Postoperative lumbar discitis, surgical vs conservative management, a retrospective study

Conclusion

Surgical management of post operative lumbar spondylodiscitis provides less rate of complications than conservative management. There is statistically significant difference between both method of management as regards of post operative complications, VAS score improvement, final clinical outcome (KWC) and post-operative neurological status according to ASIA score.

Introduction

Spondylodiscitis is an uncommon but serious infection of the intervertebral disc. the highest incidence happens between the 5th and 7th decade of life. The annual incidence varies from 0.4 to 2.5 per 100,000 people. Spondylodiscitis develops as a result of a number of factors, the most common of which being spontaneous bloodstream infections and postoperative to spine surgery.

Spondylodiscitis develops as a result of a number of factors, the most common of which being spontaneous bloodstream infections and postoperative to spine surgery. S.aureus is the most common pathogen responsible for

spondylodiscitis. due to the vague symptoms and signs, there is a large delay of 2-6 months before proper diagnosis, particularly in spontaneous type⁽¹⁾.

Individuals experience chronic back pain that intensifies at night and is frequently linked with radicular discomfort. Fever is less common and, in around one-third of patients, is linked with neurological impairments such as limb weakness, numbness, and incontinence. ESR is a sensitive but not specific marker of infection that is also used to assess therapy response. In more than 90% of cases, C-reactive protein (CRP) is also high ⁽²⁾.

Conventional spine radiographs have a modest specificity (57%) in diagnosing spondylo-discitis especially advanced cases, vertebral end plate irregularity with eventual fragmentation and decreased intervertebral disc space height. Because of its high sensitivity (96%), specificity (94%), and increased potential to offer precise anatomical infor-mation about surrounding soft tissues and epidural space, magnetic resonance imaging (MRI) is the gold standard method for diagnosing spondylodiscitis ⁽³⁾.

The conservative therapy consists of systemic antibiotics and bracing, while surgical fixation is the treatment of choice in the following situations; medical therapy has failed, there is a neurological deficit, an epidural abscess, increasing segmental kyphotic deformity or mechanical coronal or sagittal instability ⁽²⁾. The goal of surgical therapy is to thoroughly debride infective foci and decompress neurological element, repair of the deformed spine by pedicular screw posterior fixation with posterolateral fusion ⁽⁴⁾.

In cases of destructive anterior lesion of Spondylodiscitis, anterior debridement and fusion with or without posterior fixation appears logical, additional posterior stabili-zation with anterior surgery can correct and avoid kyphosis, posterior laminectomy alone is avoided as it will increase or lead to segmental kyphotic deformity ⁽²⁾, but complication rate of anterior spinal surgery is relatively higher than that of posterior spinal surgery especially when performed in patients with chronic lesions with adhesion, in thoracolumbar and lumbosacral junctions, or in patients with a history of previous anterior surgery. Combining anterior and posterior approaches may increase the risks, especially in immunocopromised patients^(4).

Despite considerably improved diagnostic possibilities, there are no standardized treat-ment guidelines for management due to a heterogeneous patient population and multiple therapeutic options ⁽⁴⁾.

One of the new classifications in spondylodiscitis diagnosis and treatment is based on three traits:

1. the inflammatory marker C-Reactive Protein (CRP) (mg/ dl),
2. Pain according to a numeric rating scale (NRS) and
3. Magnetic Resonance Imaging (MRI)
4. A Spondylodiscitis Scoring System:

SponDT Spondylodiscitis Diagnosis and Treatment

Patients and Methods

Forty cases with post-operative lumbar spondylodiscitis admitted to Minia university hospital reviewed retrospectively, twenty of them underwent surgical decompression with fixation and the other twenty were treated conservatively.  Inclusion criteria - Patients who underwent previous lumbar spine surgery, presence of neurological deficit, epidural or paravertebral abscess, progressive bony destruction with segmental deformity or instability Patient is medically fit for surgery.

Exclusion criteria

• Spontaneous spondylodiscitis
• Cervical or dorsal spondylodiscitis
• Patient is medically unfit for surgery.
• Patient refuse to do operation.

All patients were subjected to full history, general and neurological examination. VAS analogue pain score was used to evaluate pain intensity pre and postoperative while ASIA score was used to evaluate motor status. Plain x-rays anteroposterior, lateral, flexion and extension views to evaluate affected level vertebral body destruction, lumbar lordosis and kyphosis and Disc Space Height.

 Magnetic Resonance Imaging of lumber spine (MRI) was used to evaluate nature of vertebral body destruction neural elements status epidural and paravertebral involvement.

Vertebral body destruction was evaluated using grading used by Pee et al., (2008) ⁽⁵⁾:

Grade I = Isolated discitis or discitis with minor destruction of endplates in 15 cases (35%)

Grade II = Discitis with moderate endplate destruction in 20 cases (50%)

Grade III = Discitis with destruction of the VB in 5 cases (5%)

Conservative management:

Treatment consisted of immobilization with a spinal brace in 20 (50 %) patients, and antibiotics were administrated to all patients. Intravenous antibiotics were administrated until a significant decrease in C-reactive protein (CRP) was seen and for a minimum of 4 weeks.

Parenteral therapy was followed by oral antibiotics for a total of 3 months. Vancomycin and ceftriaxone were the drug of the choice for the initial empirical I.V antibiotic therapy typically for 4-6 weeks followed with oral antibiotics for additional 4-6 weeks or treatment with I.V antibiotics continued until the ESR normalizes, then changed to oral antibiotics.

Operative technique

The patients' surgical fitness was evaluated to determine whether they were candidates for surgery. psoas abscesses were drained by ultrasound guided aspiration. All patients were administered third-generation cephalosporines 2 hours before surgery. Hypotensive General anaesthesia was used to reduce intraoperative bleeding.

1-Patient positioning: The patients were put in a prone position over a spinal frame that supported their chest and pelvis, allowing the anterior abdominal wall to clear the table and reduce intraoperative haemorrhage.

2-Surgical incision: A midline posterior incision was made over the diseased vertebra or over the prior incision, the incision was prolonged proximal and distal to the anticipated level to be fused.

Bilateral subperiosteal muscle separation from the spinous process and the lamina of the affected levels till the transverse process of each level exposing the facet joint, maintaining the integrity of its capsule as much as possible.

Insertion of pedicular screws is done image guided, distraction is then done to widen the disc space. then laminectomy is done exposing the underlying dura. The inferior articular facet may be removed for good exposure of the underlying disc and Neuro-foraminotomy is done.

The disc space is then opened using a no.11 scalpel and adequate debridement of the disc material and endplate is done using disc rongeurs and curettes of different sizes. In cases where interbody cage is to be inserted, a cancellous bone graft is packed in the disc space around and inside the cage. Closure in layers is done after suction drain insertion below the lumbar fascia, the drain is removed after 24-72 hours or if it drains less than 50 ml.

Post-operative management:

Intravenous antibiotics were taken for 4-6 weeks followed by oral antibiotics according to culture and sensitivity for 4-6 weeks close and strict follow up with ESR and CRP.

Discharge: Patients were discharged from the hospital 5-7 days postoperatively and some patients from stayed up to 10 days.

Post- operative follows up:

Patients were followed up at regular interval at 2 weeks, 1 month, 3 months, 6 months and 1 year:

1- Clinical assessment:

1. a) Pain assessment (VAS for back and leg pain)
2. b) Neurological assessment ASIA motor index

2- Laboratory investigations: CRP and ESR

3- Radiological assessment:

After surgery: radiological follow up include postoperative plain x- ray

and MDCT when needed to evaluate screws, corrected deformity and late interbody fusion.

Definitive fusion: appearance of trabecular bridging pattern of dense cortical bone across graft host interface, no motion in flexion extension radiograph, no gap at interface.

Probable fusion: no trabecular bridging but no detectable motion and no identifiable gap at interface.

Functional outcome:

 Kirkaldy-Willis criteria (KWC) defined as⁽⁶⁾

(a) Excellent: Patient returned to normal work and other activities with little or no complaint.

(b) Good: Patient returned to normal work but

may have some restriction in other activities and may -on occasion after heavy work- have recurrent back pain requiring a few days’ rest.

(c) Fair: Patient work capacity was reduced, necessitating a lighter job or working part time, and may occasionally have pain recurrent requiring absence from work for one or two weeks once or twice a year.

(d) Poor: Patient does not return to work.

Results

The study involved two groups of patients each had 20 patients, the first group underwent conservative management and the other group underwent surgical fixation, 16 of them by screws only with posterolateral fusion while 4 patients by screws and interbody cage.

The age of patients ranged between 37 years and 67 years    with mean/SD (52.57±7.09). 26 patients were males (65%) and 14 were females (35%).

As regarding associated comorbidity,15 patients were diabetic (37.5%) 10 patients were hypertensive (25%), 3 were HCV positive (7.5%), 2 were Cardiac (5%) and 10 patients (25%) with no associated comorbidity.

The involved levels were (L1–2) with VB destruction in 1 patients, (L2–L3) in 7 patients, L3–4 in 4 patients, (L4–5) in 12 patients and (L5–S1) in 16 patients.

25 patients (62.5%) complained from back pain and leg pain, 15 patients (37.5%) complained from only back pain. All patients (100%) had signs of discitis in MRI findings. All conservatively treated patients had signs of discitis in MRI with no psoas abscess nor epidural collection. 4 patients only (20%) from surgically treated patients had epidural abscess that underwent to surgery immediately and 2(10%) with Psoas abscess who were referred to radiology consultant to insert flexima catheter then underwent surgical fixation.

The previous surgery for our patients was as follows; 22 (55%) patients underwent laminectomy and discectomy, 14 (35%) patients underwent laminectomy only, 2 (5%) patients underwent fenestration, while 2(5%) patients had done 2 surgeries each, the first was fenestration and the latest was a full laminectomy.

The culture in patients who were managed surgically showed that the most common organism was Staph. Aureus in 6 (30.0%), pseudomonas in 2 (10.0%), klebsiella in 1(5%), E. coli in 1 (5%) patient, Strept. Pneumoniae in 1 (5%) and there were 9 (45%) with no growth.

All patients who were managed conservatively and in surgically treated patients who had negative culture, took I.V empirical antibiotics Vancomycin and 3rd generation cephalosporin for 4 to 6 weeks then oral for the same period

In surgically treated patients with positive culture, antibiotics were given according to culture and sensitivity as follows; Penicillin or cephalosporins or vancomycin can be used for staph. Aureus infection, Combination of (penicillin or cephalosporin) and amino-glycoside were given in pseudomonas and klebseilla infected patient, (also carbapenems with anti-pseudomonal quinolons may be used in conjunction with an amino-glycoside). 3^rd generation cephalosporins in patient infected with E-Coli Penicillin and macrolides were used in patients with streptococcus pneumoniae.

The pre-operative neurologic status according to ASIA was E in all patients who managed conservatively, returned to ASIA C in one patient (5%), ASIA D in one patient (5%) and the other 18 (90%) ASIA E   after management.

In all patients who were managed surgically the pre-operative neurologic status was ASIA E in 10 patients (50%), ASIA D in 8 patients (40%), one patient (5%) ASIA C and one patient (5%) ASIA B, 18 (90%) patients returned to ASIA E, one patient (5%) to ASIA B and one patient (5%) to ASIA C at final follow up.                           

The   mean   VAS   score   of   back   pain   and   leg   pain   improved from 6.95 ± 1.32 before conservative treatment to 5.15±2.46 after conservative treatment. And improved from 7.1±1.48 pre-operatively to 3.1±1.97 post-operatively in patients who surgically managed.

Based on the Kirkaldy–Willis functional outcome criteria ⁽⁶⁾ 4 (20%), 6 (30), 2 (10%), 8 (40%) patients had excellent, good, fair and poor outcomes respectively in patients who were treated conservatively. In patients treated surgically 13 (65%), 4 (20%), 2 (10%), 1 (5%) patients had excellent, good, fair and poor outcomes respectively (Figure 1).

Regarding hospital stay 50% of patients who were managed conservatively were admitted for 5-7 days while the other 50% of the same group were admitted for 15-21 days.

While 85% of patients who were managed surgically were admitted for less than 7 days and the remaining 15% stayed for no more than 10 days.

Regarding fusion in postoperative patients, in the 16 patients who underwent posterolateral fusion combined with pedicular screws, 12 patients showed definitive fusion in the 6 months follow up. While the other four patients who underwent interbody fusion with interbody cage combined with pedicular screws showed definitive fusion in the 6 months follow up.

In patients who were managed conservatively 10 patients (50%) followed up to 6 months and they were completely relieved, and the other 10 patients (50%) developed complications that needed surgery but not included in our surgical group.

While in patients who were managed with surgical fixation; 8 patients (40%) of them were completely relieved few days post-operative, 6 patients (30%) relieved after 2 weeks, 3 patients (15%) of them were relieved after 1 month and 3 patients (15%) completely relieved after 3 months.

Case presentation:

A female patient 44 years old with previous history of L5-S1 discectomy with L5 laminectomy, presented with sever Low back pain and lt sciatica, patient was admitted for surgery, L4-5-S1 fixation with interbody cage insertion between L5-S1 was done as L4 pars showed lysis. (figures 2-4)

Discussion

The management of spinal infections remains a subject of ongoing debate, presenting both advantages and challenges for spine surgeons. The primary course of treatment is conser-vative, involving the administration of systemic antibiotics, rest, and the use of bracing⁽⁶⁾. However, surgical intervention is recommended for instances that do not show improvement with medical treatment, as well as cases presenting with neurological impairment, epidural abscess or paravertebral abscess, progressive segmental kyphotic deformity or mechanical instability, and persistent immobilizing discomfort ^(7).

In our study, we observed a higher proportion of males, accounting for 65% of the participants. This finding aligns with the research conducted by Shetty et al.,⁽⁸⁾, where 70% of individuals with spondylodiscitis were male and 30% were female. While our findings indicated a higher proportion of males, other studies, such as the one conducted by Loibel et al.,⁽⁹⁾, found roughly equal percentages of males and females. In their study of 114 patients, 55 (48%) were female.

In our investigation on patients with co-morbid conditions Out of the total of 30 patients, 75% of them had co-morbid diseases. Specifically, 15 patients, which is equivalent to 37.5%, were diagnosed with diabetes. Out of the total number of patients, 10 (25%) had hypertension, 3 (7.5%) tested positive for HCV, 2 (5%) had cardiac conditions, and 10 (25%) had no other comorbidities.

This finding is consistent with the study conducted by Loibel et al.,⁽⁹⁾, in which 88 out of 114 patients (76%) were found to have co-morbid disease. Shetty et al.,⁽⁸⁾ contradicts our data, as they found that 9 out of 27 patients (30%) in their study had co-morbid illnesses.

This disagreement can be attributed to the fact that his study focused on a small sample size of only 27 individuals, who had a low average age and did not have comorbidities such as diabetes, hepatitis C, and hypertension, which are prevalent in our country.

The most commonly afflicted lumbar levels were determined to be L4-5 and L5-S1, which is consistent with previous studies by Ha et al. (2007) and Lee & Suh ^(10,11).

The primary symptom seen in our study is back pain, which is present in all patients. Additionally, leg discomfort is reported in 25 patients (62.5%). This finding is consistent with the study conducted by Chen et al.,⁽¹²⁾, where all patients experienced back pain and 70% of them also reported leg pain.

Seventeen individuals, accounting for 85% of the total, reported back discomfort as the main symptom. Among the three surviving patients, two exhibited leg weakness without experiencing any discomfort, and one alone displayed fever. Out of the total number of patients, only six individuals, which accounts for 30% of the group, experienced fever. Out of the total number of patients, specifically eleven individuals experienced weakness in their limbs. Additionally, one of these patients suffered from a total loss of movement in their lower limbs. Out of the 11 patients, 4 reported dysethesias, and 3 experienced objective sen-sory loss. Additionally, 3 individuals exhibited acute retention of urine⁽¹³⁾.

The average blood loss for patients who had surgical fixation was 318.5±90.45 cc, with a range of 180-500cc. The operative time was 134.45±9.06 minutes, ranging from 120 to 156 minutes. The mean blood loss during surgical fixation was 430 cc, with a range of 100 to 1500 cc. The duration of the operation was 128 minutes, with a range of 60 to 240 minutes.   

All patients included in our study exhibited MRI evidence of discitis, with only 4 (10%) of patients presenting with epidural abscess and 2 patients displaying Psoas abscess. These fin-dings are consistent with the study conducted by Chenoweth et al., ⁽¹⁴⁾, which reported that 90% of patients had indications of discitis in their MRI scans, but only 10% had an epidural abscess.

Lee and Suh (2010) challenge the findings of our investigation, which reported that 10 out of 18 patients (60%) had epidural abscess.

In relation to the organism responsible for discitis, 9 patients (45%) exhibited no growth in culture, 6 patients (30%) were infected with Staphylococcus aureus, 2 patients (10%) were infected with Pseudomonas, 1 patient (5%) was infected with Klebsiella, 1 patient (5%) was infected with Escherichia coli, and 1 patient (5%) was infected with Streptococcus pneu-moniae. This finding is consistent with the study conducted by J. J. Lee et al.,⁽¹⁵⁾, which involved 18 patients. Out of these patients, 9 (50%) showed no growth in culture, 5 (27%) were diagnosed with staph aureus, and the remaining cases were identified as C. luteola, MRSA, and B. cepacia.

This contradicts the findings of Valancius et al., ⁽¹⁶⁾, whose study included only 9 patients. In their investigation, they found that out of the 3 patients with positive cultures, Staphylo-coccus aureus was present in only one patient, while two patients had Escherichia coli, two had group B streptococcus, and one had corynebacterium. The discrepancy between these findings and the current study could be attributed to the small sample size of Valancius et al.'s study.

In our study, treatment failure was observed in 10 patients (50%) who had conservative management and in 3 patients (15%) who underwent surgical fixation. Chenoweth et al., (2014) corroborates our study findings. He observed that the surgical approach to managing postoperative spondylodiscitis resulted in improved clinical outcomes and fewer complications compared to conservative therapy.

Valancius et al., ⁽¹⁶⁾ present contrasting findings to our study. In their research involving 196 patients, 75 patients underwent surgical treatment, and 31 (41%) of them experienced postoperative complications. On the other hand, among the 121 patients who received conservative treatment, complications occurred in 27 (24%) of them.

This discrepancy can be attributed to the fact that his research encompassed a sample size of 219 patients spanning from 2000 to 2010. His findings indicated that conservative treatment is a secure option for patients who are carefully selected and do not exhibit any spondylodi-scitic complications. Additionally, the favor-able hospital environment and effective infection control measures contribute to succe-ssful conservative management of patients.

In our study, all patients who were managed conservatively had a neurologic status of E according to ASIA. Out of these patients, one (5%) progressed to ASIA C, one (5%) progressed to ASIA D, and the remaining 18 (90%) remained at ASIA E after management.

The pre-operative neurologic status of all surgically managed patients was as follows: 50% (n=10) had ASIA E, 40% (n=8) had ASIA D, 5% (n=1) had ASIA C, and 5% (n=1) had ASIA B. At the final follow-up, 90% (n=18) of patients returned to ASIA E, 5% (n=1) improved to ASIA B, and 5% (n=1) improved to ASIA C.

This is consistent with the findings of Lee et al., ⁽¹⁵⁾. Within his analysis of 18 patients, 7 individuals progressed from ASIA D to ASIA E, 5 patients advanced from ASIA C to ASIA E, and 3 patients reverted from ASIA B to ASIA E after the surgical procedure.

Furthermore, this aligns with the findings of Shetty et al.'s study ⁽⁸⁾, which reported that out of their 5 patients with neurological abnormalities, ASIA D improved to ASIA E post-operatively, and one patient improved from ASIA C to ASIA D.

In our study on the VAS score for back and leg pain before and after management, we observed.  improvement in the conservative group was as follows; the VAS score decreased from 7 before treatment to 4.5 after therapy. The surgical group experienced a statistically significant improvement in VAS score, decreasing from 7 to 2 postoperatively.

The surgical group demonstrated a more substantial improvement in VAS score compared to the conservative group, and this difference is statistically significant. These findings corroborate Epstein's assertion (Epstein, 2015) that the VAS score in his study shown improvement following surgical intervention. Consequently, these results provide more evidence that surgical care yields a favorable outcome in terms of quality of life.

Based on the final clinical outcome, our study found that after surgical intervention, 65% of patients were classified as excellent, 20% as good, 10% as fair, and 5% as having a poor outcome. After conservative management, 20% of patients were excellent, 30% were good, 10% were fair, and 40% had a poor outcome.

These findings are consistent with the study conducted by Shetty et al., ⁽⁸⁾, which reported that out of a total of 27 patients, 14 (52%) had outstanding clinical outcomes following surgical intervention. Additionally, 9 (33%) patients had good outcomes, 3 (11%) had fair outcomes, and only one patient (3%) had a bad clinical outcome.

This contradicts Epstein's study, since only 3 out of 9 patients (33%) achieved an excellent clinical outcome, while another 3 patients (33%) had a good clinical outcome, and 2 patients (22%) had a bad clinical outcome. This discrepancy may be attributed to the small sample size of just 9 patients in Epstein's study.

In our study, the length of hospital stays for patients who received conservative manage-ment ranged from 5 to 7 days for some patients and 15 to 21 days for others. For patients who underwent surgical management, the majority had a hospital stay of 2 to 3 days. There is a statistically significant disparity in hospital stay duration between patients who underwent surgical care.

These findings align with the results of Loibl et al's study⁽⁹⁾, which showed that patients who had conservative management had a median length of stay of 29 days, whereas those who underwent surgical management had a median length of stay of 9 days.

This contradicts the findings of Valancius et al.,⁽¹⁶⁾ The patients who received conservative treatment were hospitalized for an average of 20 days, whereas those who underwent surgery stayed in the hospital for an average of 25 days. This difference can be attributed to the fact that the study included 219 patients, out of which 117 underwent surgery, and different surgical techniques were used.