Intra-discal drain insertion for culture and drainage of pyogenic spondylodiscitis: A one-step diagnostic and therapeutic procedure

Abstract

Background

Pyogenic spondylodiscitis is diagnosed in recent years at higher rates due to the aging population, increased survival of chronic and immune suppressed patients, and the higher rate of invasive procedures leading to bacterial seeding or direct contamination of the disc space. Treatment guidelines encourage bacterial sampling before initiation of antibiotic therapy, and drainage of pus collections. We present our experience with percutaneous CT-guided drain insertion into the disc space itself as a one-step procedure for both culturing and subsequent continuous drainage of the infected disc space.

Materials and methods

We retrospectively reviewed all cases of pyogenic spondylodiscitis admitted to our spine surgery unit during the past five years and treated with CT-guided percutaneous drain insertion into the infected disc space. All patients were followed until complete resolution of the infection.

Results

We retrieved electronic records of 12 patients, none presenting with neurological compression symptoms. Cultures taken at the time of drain insertion were positive in 10 patients (83.3%), much higher than the reported yield for needle aspiration (14–48%) and comparable to the yield of open biopsy. In all patients complete resolution of the infection was reached, determined by clinical, laboratory, and imaging parameters.

Conclusions

Our retrospective case series demonstrates the feasibility and effectiveness of intra-discal CT-guided drainage of an infected disc space. The procedure does not add much burden to current practice as disc-space sampling for culture is commonly performed anyway, and adds the benefit of direct drainage of the pus at its source.

Introduction

Pyogenic discitis, the bacterial infection of the intervertebral disc, is diagnosed in recent years at a much higher rate than in the past.^1–3 This is attributed to the increase in life expectancy in the elderly, greater survival of immune compromised patients, improved and accessible imaging, and a higher index of suspicion amongst treating clinicians.^2,4,5

The disc space can be infected through hematogenous spread from a distant site, direct extension of a proximal nonspinal infection, or introduced iatrogenically during various spinal procedures.^2,3 In most cases, the infective organism is Staphylococcus aureus.^6,7

In the disc space, a relatively avascular structure in adults, the infective agent can multiply, protected from the host immune system and blood-borne antibiotics.³ Left untreated, the infection can propagate to surrounding anatomical structures, most commonly the adjoining vertebrae causing osteomyelitis, or the spinal canal causing epidural abscess, or the adjoining soft tissue—causing psoas abscess, lung empyema, or retropharangeal abscess, depending on location.²

Treatment of discitis can vary widely, ranging from antibiotic therapy alone or in combination with percutaneous or open abscess drainage, bracing, spinal debridement, corpectomy, and/or instrumentation.^3,8–10

Before initiation of antibiotic therapy, if the causative pathogen is not identified through cultures obtained otherwise, treatment guidelines recommend performing a biopsy,⁶ although the yield is reported to be very low, ranging from 14% to 48% for image-guided needle biopsy and 76% for open biopsy.^11–13

CT or fluoroscopy-guided drain insertion into the disc space is an attractive option as it allows both sampling for culture and continuous drainage of pus from the infection source itself in a single procedure. In this retrospective case series, we describe our experience with this less-established technique.

Materials and methods

Using our hospital’s electronic health records and picture archiving and communication system (PACS), we retrospectively reviewed all cases of pyogenic spondylodiscitis admitted to our spine surgery unit during the past five years and treated with percutaneous drain insertion into the infected disc space.

Included were patients in which the drain insertion was done under CT guidance and followed until complete resolution of clinical, laboratory, and radiographic evidence of disease.

Excluded were patients in whom the intra-discal drain insertion was done under image intensifier control or during an open surgical procedure.

The surgical technique for CT guided insertion was as follows:

Informed consent was obtained prior to pre-procedure check list verification and before the patient was admitted to the CT suite.

All patients underwent insertion of a percutaneous drain under CT guidance (Phillips brilliance 64 slice, Cleveland, USA) using either a Dawson-Mueller multipurpose drainage catheter 8.5/10.2 Fr (Cook Medical, USA) or an Expel drainage catheter 8.3/10.3 Fr (Boston scientific, Costa Rica).

The patient was positioned prone or oblique on the CT table depending on the expected angle of approach to the target disc space.

In lumbar discs, the insertion point and trajectory were at the disc level on the coronal and sagittal planes, and angulated about 45° on the axial plane. This trajectory ensures the spinal nerves are not injured, as they exit the neuroforamina just caudal to the pedicle and travel obliquely antero-laterally within the psoas muscle, clearing the disc space level.

In the thoracic spine, the trajectory was calculated to clear the rib head, which articulates with both vertebrae above and below the target disc. In these cases, it is not possible to ensure avoiding the exiting nerve root as it becomes an intercostal nerve, but theoretically, even if injured, the resulting neurological loss would be minor.

After an initial scan for localization with a longitudinal marker stuck to the skin, the final approach was determined and the point of skin entry marked. Almost all procedures were carried out under local anesthesia, occasionally supplemented with conscious sedation and rarely under general anesthesia, the last two provided by an anesthetist. Following antiseptic skin preparation and draping, up to 20 mL of lidocaine 1% local anesthesia (Esracaine, Rafa, Israel) was injected subcutaneously and along the approach track.

Following an approximately 5 mm skin incision, a 21G needle was inserted into the disk with or without intermittent CT guidance as required. Following appropriate positioning any fluid that could be aspirated was sent for culture. Using standard Seldinger technique (with the set wire, if provided in the set, or a separate 80 cm wire if not), the access was enlarged and the drain was positioned with the pigtail portion positioned within the disc. If no fluid could be aspirated from the drain it was flushed with 1–2 cc sterile saline and the aspirate sent for culture.

Following skin fixation and dressing, the patient was then removed from the CT suite and returned to the ward.

Vital signs were monitored overnight, and peripheral blood hemoglobin levels were checked 6 h post-procedure and the following morning to detect occult bleeding.

The catheters’ exit site was enclosed in a clear colostomy bag pouch and skin barrier (Figure 6). Daily replacement of the bag enabled discharge to be noted. The catheter was removed when discharge ceased.

Figure 6.

Patient B, drain tip in colostomy bag showing daily secreted pus.

Data collected from our hospital’s electronic patient files and PACS were: demographics, comorbidities, culture results, daily oral temperature, complete blood count (CBC), C-reactive protein (CRP) during hospitalization and ambulatory follow-up, erythrocyte sedimentation rate (ESR), numeric pain scores (NPS), radiographs, CT and MRI scans done during hospitalization and post discharge.

Statistical analysis calculated mean, median, and standard deviation for retrieved data. For time-dependent paired value changes, the Wilcoxon signed rank test was used.

This study was approved by our institutional review board. As this study reviews the results of the usual treatment, no informed consent was required.

No funds were received in conjunction with this study by any of the authors.

Results

Records of 12 patients fitting our inclusion criteria were retrieved. Excluded were patients in whom the spinal infection was treated without intradiscal drain insertion or the drain was inserted without CT guidance.

There were six males and six females aging 36 to 80 years old (average 57.2, median 57.5). None of the patients presented with neurological compression symptoms.

The drain was inserted, on average, five days following MRI diagnosis(mean 5.08, SD 4.44, median 4), and retained, on average, 20 days (mean 19.67, SD 16.42, median 18).

After initial drainage at placement, the daily drain secretion was persistent but small in volume, and noted as “smears in the bag” since measurement of the exact volume proved to be impractical.

In two of the patients antibiotics were started before drain insertion, based on peripheral blood cultures or empirically. Cultures taken at the time of insertion were positive in 10 patients (83.3%) with identification of six Staphylococcus Aureus, one Escherichia coli, one Enterobacter cloacae, one Candida albicans, one Brucella. In two patients, both cultures and subsequent polymerase chain reaction (PCR) assay were negative. They were treated empirically with vancomycin and ciprofloxacin.

The mean numeric pain score (NPS, 0–10) on admission was 4.45 (median 5, SD 1.92) and on discharge 1.73 (median 2, SD 1.74, Wilcoxon Z 2.66, p < 0.01). The mean length of hospital stay was 31.5 days (median 29.5, SD 15.22). Following discharge, all patients continued antibiotic therapy and were followed at our outpatient spine clinic until complete resolution of symptoms, normalization of CRP values and interbody spine fusion judged by standing radiographs was reached (for images of two example cases, please refer to Figures 1 to 6)

Figure 2.

Patient A, axial CT taken at drain insertion into the disc space (day 11).

Figure 3.

Patient A, final fat suppression MR demonstrating complete interbody fusion (day 354).

Figure 4.

Patient B, CT drain insertion post percutaneous instrumentation.

Figure 5.

Patient B, standing lateral radiograph with drain in place.

Figure 1.

Patient A, initial pre-treatment fat suppression MR demonstrating discitis at the T10-11 level (day 0).

Two procedure-related complications were found. One patient, due to a sharp drop in serum hemoglobin levels, was found to have a pseudoaneurysm of the insertion-side L5 segmental artery, which was embolized and a new drain inserted. The reason for this complication is most probably a trajectory miscalculation. In another patient the catheter tip broke during insertion, requiring conversion to open surgery for retained fragment removal and drainage. The reason for this complication was re-insertion of the stylet/stiffening cannula when the catheter was already partially in place, damaging its integrity and causing it to break when extraction was attempted. This patient was tabulated for complications but not for other outcomes, making the total complication rate 2/13 (15.3%). Drain removal was done when the daily secretion ceased and was uneventful in all patients. For detailed data and comorbidities, please see Table 1.

Table 1.

Patient data table.

Case number	Male/ Female	Age (years)	Comorbidities	Infected disc level	Pre NPS	Post NPS	Drain days	Pathogen	Radiographic findings
									Discitis	Osteomyelitis	Vertebral collapse
1	M	57	Diabetes mellitus, hypertension, dyslipidemia	L2-3	6	0	53	E. coli	YES	YES	YES
2	F	80	Hypertension, dyslipidemia	L1-2	5	2	33	MSSA	YES	YES	NO
3	M	45	Drug addiction, HCV positive	L3-4	5	1	42	1. Staph epidermidis 2. MRSA	YES	NO	NO
4	F	38	Arthrogryphosis, post spine surgery	L1-2	0	0	7	MRSA	YES	YES	NO
5	F	75	Pulmonary adenocarcinoma, hemodialysis, hyperlipidemia, asthma, diabetes mellitus	T9-10	3	1	20	MRSA	YES	YES	YES
6	F	75	Diabetes mellitus, chronic heart failure, dyslipidemia, chronic renal failure, hypertension	T6-7	5	0	27	Enterobacter	YES	YES	YES
7	M	72	Hypertension, dyslipidemia, diabetes mellitus	L4-5	5	0	2	Candida albicans	YES	YES	NO
8	M	47	Drug abuse	T5-6-7	4	2	1	MSSA	YES	YES	YES
9	F	67	Diabetes mellitus, hypertension, dyslipidemia, ischemic heart disease	L3-4	5	0	18	Negative	YES	YES	NO
10	M	58	None	L4-5	1	2	8	Brucella	YES	YES	NO
11	F	36	None	L3-4	4	1	18	MSSA	YES	YES	NO
12	M	37	None	T12-L1	5	1	7	Negative	YES	YES	NO

Discussion

Treatment of vertebral discitis-osteomyelitis is varied, and tailored to the wide spectrum of clinical presentations. It can range from antimicrobial therapy alone to emergent extensive surgery, with a spectrum of intermediate options.

We present a small case series in which a drain was placed under CT guidance into the disc space itself, the logic being that the infected disc space is the source of pus formation and subsequent propagation to adjacent structures, thus draining the “feeding tap” rather than the “collected pool” can theoretically prevent formation of a psoas abscess or epidural extension.

Our results suggest this technique is possible and effective in obtaining cultures and controlling infection dissemination no less than other, more common, catheter tip placements such as in a psoas abscess or the retroperitoneal space.

Our yield in obtaining positive cultures is much larger than reported for percutaneous biopsy and matches the yield reported for open biopsy (83% vs. 14–48%).^13–15 This can be explained by the greater sampling volume collected by draining actual pus from the disc space instead of the smaller volume retrieved by needle biopsy. Another explanation might be that appearance of pus from the drain serves as a verification method for correct tip placement, not possible in a “single shot” biopsy.

Another advantage of intra-discal drain placement is that daily discharge from the disc space can be monitored, serving as a clinical indicator of resolution of the infection. Typically, the drain will secrete bloody pus during the first day or two, followed by gradual transition to a clearer fluid in decreasing volumes. When the discharge ceases, peripheral blood CRP levels normalize and radiographic healing is apparent, it can be assumed that the infection has resolved.

The procedure was well tolerated by patients, with a statistically significant rapid decrease in self-reported pain scores. This in itself is beneficial, allowing patients to ambulate without the need for external bracing.

We had one technical failure, requiring open decompression and removal of a catheter tip from the disc space, and one case of inadvertent segmental artery injury requiring embolization. The high complication rate (15.3%) probably reflects a learning curve rather than inherent procedure-related pitfalls.

To avoid tip breakage, adherence to the catheter’s manufacturer instructions should be stressed. Another way to avoid insertion problems initially is to select patients with a larger disc space which is easier to penetrate even if the trajectory is not optimal. Other than trying not to stray from the disc-space level, we have no fail-safe method of avoiding a segmental or aberrant artery. As spinal segmental arteries can be recognized by a pre-interventional contrast enhanced 3D-CT, this might aid avoiding vessel injury. As a routine, we keep the patients hospitalized at least over one night and perform 6- and 12-hour post-procedure hemoglobin level testing to detect occult bleeding. These two complications demonstrate that the procedure should probably be reserved to institutions where overnight observation and inter-departmental support is available.

Our small series does not enable reaching conclusions regarding the patient population most likely to benefit from direct disc space drainage. Many patients can be treated with antibiotics alone, and in cases where a growing epidural collection is evident, open decompression of the spinal canal is warranted, at times with surgical debridement and spine instrumentation. Still, in this wide spectrum of options, we feel that when intra-discal fluid is clearly evident in imaging, drainage of the disc itself might reduce bacterial load and prevent extension into surrounding tissues, and should be considered.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.