A large proportion of patients with Low Back Pain (LBP) (80 %) are classified as "non-specific", due to the lack of a patho-anatomical diagnosis. This term is not satisfactory, since it implies uncertainty for both the patient and the clinician. Modic changes have recently been identified as a pain giving patho-anatomical diagnosis for a considerable percentage (20-30 %) of LBP patients. The causes of these Modic changes have not been determined conclusively but bacterial infection is among the plausible causes for those following a disc herniation.

Purpose The aim of this study is to assess the clinical effect of antibiotic treatment compared to placebo, in patients with LBP and Modic changes following a lumbar herniated disc.

Design A double blind prospective clinically controlled trial. Patients will be randomized by a blinded observer to either Group A: 1 antibiotic tablet 45 patients Group B: 1 placebo tablet 36 patients Group C: 2 antibiotic tablets 45 patients Group D: 2 placebo tablets 36 patients

The patients will take the drug three times a day for 100 days.

Inclusion Criteria:

• Age between 18-65 years
• Previous lumbar herniated disc confirmed by MRI
• Time since lumbar disc herniation more than 6 months less than 2 years
• LBP in the area of L1 to L5, with an added Low Back Pain Rating Scale of 6 or more
• The present MRI shows Modic changes
• Positive Springing test (SPT) at the level of Modic changes

Exclusion Criteria:

• Allergy against antibiotics
• The Beck Depression Inventory >7,
• Pregnancy
• Breast feeding
• All Kidney diseases
• Pending workers litigation or pension
• Danish as second language
• Fertile women who do not user safe anti conseption
• Reduced liver functions
• Reduced liver function due to previous treatment wiht amoxicillin/clavulanasid.