The cocktail infection: anaerobic and aerobic co-Infections in tubercular vertebral osteomyelitis [version 2; peer review: 1 approved with reservations, 1 not approved]

Background: Anaerobic organisms have been known to have an association with dental infections, bacteremia, endocarditis and soft tissue infections. However, anaerobic isolation from bone and joint infections are relatively rare. Sparse literature has been found on reports of anaerobic osteomyelitis. There is no literature reported on anaerobic osteomyelitis complicating Tuberculosis of spine. Case Report: We report two cases of tuberculosis of spine complicated by aerobic and anerobic infections. The first is a case of a young female who presented with chronic lower backache and fever. Examination revealed a lumbar scoliosis with paraspinal tenderness. Magnetic resonance imaging (MRI) of the spine showed lumbar spondylodiscitis with multiple abscesses. There were air-fluid levels noted in the abscesses. The pus sent for CBNAAT (cartridge based nucleic acid amplification test) was positive. Further the cultures also grew Escherichia coli (aerobic) Bacteroids fragilis and Peptoniphilus asachrolyticus (anaerobic) organisms. She improved with a course of intravenous


Introduction
Anaerobic organisms have been known to have an association with dental infections, bacteremia, endocarditis and soft tissue infections. 1 However, anaerobic isolation from bone and joint infections are relatively rare. This could be attributed to the fastidious nature of the organism which makes isolation from the bone and joint infections cumbersome. 1 The most common organisms implicated in anaerobic osteomyelitis are Bacteroides, Fusobacterium, Peptostreptococcus and Clostridium species. Predisposing factors include children, contiguous spread of infection, vascular disease and complicated fractures. 2 Sparse literature has been found on reports of anaerobic osteomyelitis. There is no literature reported on anaerobic osteomyelitis complicating Tuberculosis of spine/Tubercular vertebral osteomyelitis (Pott's spine). However, our cases highlight the rare isolation of gram-negative anaerobes (Prevotella, Bacteroides, Peptoniphilus) from Tubercular vertebral osteomyelitis (TVO) in immunocompetent adults diagnosed with tuberculosis of spine.

Materials/Methods
Two cases of infectious spondylodiscitis have been presented in this case series.

Consent
Written informed consent for publication of their clinical details and/or clinical images was obtained from the patients.
Case report Case 1 A 22-year-old poorly built (body mass index: 16 kg/m 2 ) lady from rural South India who was pursuing her undergraduate degree, presented with complaints of low back ache of one-month duration. It was insidious in onset and gradually radiated to the right gluteal region. She was able to walk only a few steps with a limp. There was no bowel or bladder disturbance. She preferred to keep her right hip flexed to about 30 degrees in supine position. The backache was associated with high grade fever with chills and rigors. There was no history of tuberculosis in her family or her community. She had no significant history suggestive of immunosuppression. She had no known co-morbidities. There was no significant genetic history given by the patient.
On examination, she was febrile (102°F) with a pulse rate of 98 beats/minute. There was midline and right paraspinal tenderness from L3 to sacral region. Neurological examination of lower limbs was unremarkable. There was no focal neurological deficit.
Lab investigations showed that she was anaemic (Haemoglobin of 8 g/dl). Her total counts were 13600/mm 3 (predominantly neutrophilia). The ESR was elevated (88 mm/hour) and CRP (261mg/L). The renal function tests were within

REVISED Amendments from Version 1
The manuscript has been revised to include additional details of the patients' physical examination, chest radiograph findings, MRI images, details of treatment given and outcome.
In case 1, we have described that the patient was febrile on admission with paraspinal tenderness in the lumbo-sacral region. Chest Radiograph was done and found to be normal.
Additional radiological images 1c and 1d have been included for case 1 to indicate spondylodiscitis.  Follow-up details and outcome also included.
In case 2, we have confirmed that there were no co-morbidities or any relevant psych-social or genetic history obtained from the patient. We have also noted that the patient had an improvement in the anemia, ESR and weight gain on follow-up.
Literature search to find similar cases with tubercular vertebral osteomyelitis and anaerobic infection was also done and details included in the manuscript.
The strengths and limitations of this case have also been detailed after the discussion.
Any further responses from the reviewers can be found at the end of the article normal limits, while serum albumin was only 1.9 mg/dl. Blood and urine cultures were sterile. Her chest x-ray showed no evidence of tuberculosis.
A radiograph revealed mild lumbar scoliosis to the left side and bulky psoas muscle in the right side (Figure 1a and b).
MRI spine revealed features suggestive of spondylodiscitis of L5-S1, spondylitis of L4 with abscesses adjacent to L4, L5, S1, S2 and S3 vertebra in the right side. There was also air-fluid level in the abscess (Figure 1c and d). The spondylodiscitis can be appreciated in the Figure 1e which shows post-contrast enhancement in L5-S1 vertebral discs. Figure 1f shows the radiographs before treatment and after treatment for comparison.
In view of the psoas abscess and radiological evidence of spondylodiscitis, she underwent surgery involving decompression of L5 vertebra and exploration of L5-S1 space with drainage of the abscess. Intra-operatively foul-smelling gas escaped from the abscess. The L5-S1 disc also had purulent collection. Pus was sent for CBNAAT (cartridge based nucleic acid amplification test) for tuberculosis, aerobic and anaerobic cultures. Material from L5-S1 disc was sent for histopathology.
On the third hospital day CBNAAT report was suggestive of Mycobacterium tuberculosis she was started on anti-tubercular therapy (ATT). The next day the aerobic culture grew Escherichia coli whereas anaerobic culture grew Bacteroids fragilis and Peptoniphilus asachrolyticus. In addition, she was also started on injection Piperazillin-Tazobactem and oral tinidazole for 10 days. There was wound dehiscence on tenth postoperative day.
The repeat culture from the pus drained grew only E. coli with similar sensitivity though the growth was scanty. We continued antitubercular drugs and added Injection Amikacin and Imipenem (as per the culture sensitivity report) for a total of three weeks. She was taken up for secondary suturing during this period. Patient was discharged subsequently. On follow-up at one month, her symptoms had improved and her anemia had improved to 9.5 gm%. She had also gained 1.6 kgs in weight. Repeat radiographs and MRI were done at the three-month follow-up and showed resolution of the abscesses. Patient was adherent and tolerated the anti-tubercular medications well. She was continued on ATT for 18 months. She did not have any ATT related complications and was compliant with the treatment throughout the follow-up.

Case 2
A 54-year-old man from an urban district of South India, a tailor by occupation, came with complaints of insidious onset of neck pain of about three weeks duration. The pain was gradually progressive over the past 10 days. The pain would aggravate on movements of the neck, but there was no radiculopathy. There was also associated fever with chills and rigors. He reported significant weight loss in the recent past. There was no history of medical co-morbidites. There was no relevant psycho-social or genetic history obtained from the patient.
On examination, neck appeared short and kyphotic. There was midline tenderness from C4-C7 region. Movements of the neck was grossly restricted due to pain and spasm. He was unable to flex his neck without significant pain. The elbow flexion was MRC Grade 4/5 on both sides. There was no sensory deficits and reflexes were normal. He was anaemic (Hb 10 g/dl) with neutrophilia. ESR was moderately elevated at 68 mm/hour. His chest radiograph was normal.
In an X-ray of the cervical spine, we noted increased prevertebral soft tissue shadow and erosion of vertebral end plates at C4-5 level with gross kyphosis. Air-fluid levels were also seen in the prevertebral soft tissue shadow (Figure 2a and b).
MRI of the cervical spine confirmed spondylodiscitis of C4-C5 with cord compression. Multiple collections in parapharyngeal space communicating with the pre-vertebral space were also seen along with air-fluid levels (Figure 2c and d) By this time the blood culture grew Streptococcus constellatus and he was started on Injection ceftriaxone (2gm q24h) as per sensitivity report. His urine culture was sterile and Brucella agglutination test was negative.
He underwent anterior cervical decompression, by C5 corpectomy and fusion (Figure 2e and f). A distinct foul-smelling odour was noted intra-operatively and around 100 ml of pus was drained. The C4 and C5 vertebrae were completely destroyed with surrounding caseous material. Pus was sent for aerobic and anaerobic cultures and the caseating tissue was sent for histopathology. Patient tolerated the procedure well.
The aerobic bacterial culture from the pus grew Streptococcus constellatus with similar sensitivity as that of his blood culture. Anaerobic culture grew Prevotella sps. The CBNAAT report was positive for Mycobacterium tuberculosis.

Histopathology confirmed granulation tissue with epitheloid cells and Langhans giant cells suggestive of tubercular osteomyelitis.
He was started on Antitubercular therapy and Injection Metronidazole (500 mg q8h for two weeks) in addition to Injection Ceftriaxone (two weeks). The wound healed well. Anti-tubercular treatment was initiated and the patient was followed up on OPD basis. He was compliant with treatment and took the ATT for a total of 15 months. Improvement in anemia and ESR, weight gain as well as radiological improvement was noted in this patient on follow-up.

Discussion
Anaerobic osteomyelitis was first reported in 1884 by Von Langenbeck in case of vertebral osteomyelitis. 3 Decades later, Taylor and Davies noted the presence of anaerobic organisms within sequestra in 55% of patients with chronic osteomyelitis. 4 It was found that these anaerobic organisms were more frequently isolated from the inside of the sequestra and were usually mixed isolates i.e two or more anaerobic isolates were isolated. The chronicity of the osteomyelitis was directly proportional to the frequency of anerobic isolates found. In a retrospective study done on osteomyelitis by Lewis et al., it was found that 39 percent of patients with osteomyelitis had anaerobic infections. 1 This serves to prove that anaerobes play a much larger role in osteomyelitis than known previously. It is likely that these infections are less reported due to poor awareness of their prevalence and the cumbersome methods of isolation.
However, these organisms were found as a part of the mixed flora which included gram positive cocci, gram negative bacilli and other anaerobes. This is consistent with the pattern of mixed infections reported in our case report. Our first patient was found to have anaerobic Prevotella infection from epidural abscess with associated Streptococcus bacteremia in a chronic tubercular vertebral osteomyelitis. Our second case was also found to have a combination of anaerobes comprising of Bacteroides, Peptoniphilus (Peptotreptococcus) and Clostridium with gram-negative Pseudomonas aeruginosa isolated from a psoas abscess secondary to a chronic tubercular vertebral osteomyelitis.
The predisposing factors contributing to anaerobic osteomyelitis usually include children, diabetes mellitus, oral infections or procedures, upper respiratory tract infections, trauma, peripheral neuropathy and complicated fractures. 2 It is interesting to note that both our patients had no pre morbidities to predispose them to anaerobic osteomyelitis.
In a review by Raff and Melo of a large series of 193 anaerobic osteomyeltits cases collected from the world literature published between 1936 to 1976, it was the found that the most common anaerobe implicated in osteomyelitis was Bacteroides followed by Peptostreptococcus and Fusobacterium. 5 This trend was confirmed by a study by Ziment et al. who found the most common isolate being Bacteroides followed by Peptostreptococcus and Fusobacterium. 6 This similar trend of occurrence was noted in a large study done on 134 cases of pyogenic osteomyelitis by Haider et al. between 1992Haider et al. between to 1993 This trend agrees with our case of anaerobic osteomyelitis which showed a mixed anaerobic isolate of Bacteroides with Peptostreptococcus and Clostridium.
A review of literature on available data on anaerobic osteomyelitis with Prevotella isolates shows only nine cases reported till date. [8][9][10][11][12][13][14][15][16] Among this only one case reports secondary Prevotella infection on chronic tubercular vertebral osteomyelitis. A summary of the data available on Anaerobic osteomyelitis with prevotella isolates are shown in Table 1.
A literature search was done to identify other published reports of anaerobic infections complicating tubercular vertebral osteomyelitis. However, we found only one case report by Salavert M et al who reported a combination of Prevotella and Candida infection in an intravenous drug abuser diagnosed with tubercular vertebral osteomyelitis in 1997. 17 Both our cases did not show any evidence of hematogenous osteomyelitis. This finding concurs with evidence found by in a review study on 61 cases of anaerobic osteomyelitis of long bones 18 where 29.5% of the cases were hematogenous and 32.8% were exogenous osteomyelitis. The study by Lewis et al. also shows a preponderance towards exogenous osteomyelitis. 1 In both our cases we could not ascertain how these polyorganisms could infect the spine. Management of anaerobic osteomyelitis includes a two-pronged approach with adequate drainage of purulent material and parenteral administration of antibiotics for at least four to eight weeks. 19 The antibiotics were decided according to the sensitivity pattern noted on culture.
Strengths and limitations: The strength of this case series is the similar clinical presentation of both the cases and the presence of aerobic and anaerobic infections in both of them. However, we were unable to follow up the second case beyond two months as he opted for care under a local orthopaedician due to travel constraints. This was one of the limitations of our case series. As a result of this adherence and tolerance of the treatment could not be assessed in the second case.

Conclusion
Though tubercular vertebral osteomyelitis is usually a diagnosis in itself, it should not hinder us from considering secondary infections (both aerobic and anaerobic) complicating the osteomyelitis. Further, the presence of air-fluid levels on imaging studies and the presence of foul smell during operative exploration of the spine must arouse the suspicion of an anaerobic co-infection. Isolation and treatment of these organisms are crucial as they may hamper the clinical outcome of the primary TVO. Therefore, the cocktail of organisms, both aerobic and anaerobic isolated in these cases highlights the importance of suspecting, isolating and treating these lurking organisms in cases of tuberculosis of spine.
Patient perspective 1. I was suffering from severe back ache for over 2 years with recurrent fever. I was unable to continue my studies also as sitting in class was very painful. I received treatment at several local hospitals but the fever persisted and the pain was not alleviated. Following admission at Kasturba hospital, Manipal, I underwent an MRI of my spine and the doctors found the source of infection in my spine. I underwent a decompression surgery with treatment with TB medicines and intravenous medications at the hospital. The fever subsided in the first week following treatment. Gradually I was able to move around with support and at my 3 month follow-up, I walked without pain to the hospital. I am much better now and have rejoined college to continue my studies.
2. I had neck stiffness and pain for many years. The pain radiated to my arms and I struggled to sit and work for long hours with my neck bent forward. In the hospital, a team of doctors examined me and suspected a spine infection. An MRI confirmed infection of the neck bones and pus collections. I underwent a surgery on my neck and received a long course of medicines in the hospital and at home. I noted a significant improvement in my pain and fever. I was symptom free by my 2-month follow-up at the hospital.