| Methods | Results | The Theory | My Comments | Our Future Project |

Albert HB, Sorensen JS, et al. Antibiotic Treatment in Patients with Chronic Low Back Pain and Vertebral Bone Edema (Modic Type I Changes):a Double-Blind Randomized Clinical Controlled Trial of Efficacy. Eur Spine J 2013;22:697-707

In 2013, Albert et al. startled the world of spine research by publishing the results of a double-blind randomized placebo-controlled trial (the highest quality of research there is) which indicated that a simple course of antibiotics was highly effective at reducing chronic back and leg pain in a group of patients who had failed operative care and/or conservative care (a tough group of patients to treat). I agree with his recommendation that patients with disc herniation and Modic change type I at the same level, and who have failed conservative/operative care, may consider a 100-day course of antibiotic therapy in hopes of getting rid of that pain.

Methods: Only patients who had MRI-confirmed disc herniation, Modic type I change at the same level, and had failed at least six months of conservative care or operative care (discectomy) were enrolled into the study, at which time a new MRI was performed to confirm the volume of Modic type I change. By computer assignment, qualifying patients were given either the real antibiotics (amocicillin-clavulanate Bioclavid™, which is an antibiotic that has low anti-inflammatory properties and does not penetrate the disc) or a fake antibiotic (placebo made of calcium carbonate, but looked identical to the real thing) and asked to take these for 100 days (they were given a medication diary to ensure they complied), which is a time-period commonly employed on patients with confirmed discitis. 90 patients started the real antibiotics and 72 started the placebos. (The randomization was actually a little more complicated, but this accurate and is good enough for our purposes). They were discouraged from doing any other type of treatment or even consultations with medical specialists for the next year; however, they were allowed to continue taking whatever anti-inflammatory/pain medication they were taking prior to the study. Although they used some strange and uncommon outcome assessment tools, they did include the Roland-Morris (a disease specific disability tool) and a simple 0-10 numeric rating scale for back and leg pain. Blood serum from the patients was also taken before the start of the study and at the one-year time-point.

Results: of the 160 patients entered the study, 144 (88.9%) completed follow-up that one year, which included a repeat MRI. Patients in the antibiotic group achieved higher improvement on all outcome measures as compared to the placebo group (see table below), this included leg pain. One weird thing was that this improvement did not start for most patients until 6-8 weeks after the start of the real antibiotics – that's a long time. The volume of Modic type I change also decreased in antibiotic group versus placebo group; however, the placebo group just happened to have more patients with small volume of Modic change at the start of the study, so I'm not impressed and this result should be definitely taken with a grain of salt.

23% of patients in the antibiotic group did consult a medical specialist (we don't know what type of treatment they got); however, 42% of the placebo group sought extra help from medical specialists. Therefore, the improvement in antibiotic group cannot be attributed to extra help from a medical specialist.
Conclusions: the investigators concluded that antibiotic therapy should be considered, not for all patients who suffer low back pain, but for patients who have failed conservative or operative care for chronic back/leg pain and have a history of disc herniation with Modic type I change.

The theory: Albert's postulates that common skin bacteria can enter the bloodstream through brushing your teeth/flossing and then entered the disc via the new blood vessels that form around a disc herniation and down the annular tear. Most of the bacteria found in all the other studies are anaerobic gram-positive, which means they live only in a low oxygen environment, such as the inside of the disc which normally have no blood muscles. Once the bacteria (especially one called p. Acnes, which is the same one that causes acne) take up residency inside the disc, they secrete, as part of their normal metabolism, an acid (propionic acid) as well as tumor necrosis factor alpha (an evil proinflammatory mediator [a fire starter]), which diffuse from the disc into the vertebral endplates and cause destruction of bone and marrow which in turn starts a general inflammatory process which in turn sets off the pain-carrying nerve fiber in the vertebral employing and causes low back pain. This process shows up as Modic type I change on MRI.  The relief of the leg pain, however, was surprising. The authors theorized that the acid and tumor necrosis factor alpha, generated by the bacteria, may have leaked out of the disc onto the nerve roots and in turn caused the patient sciatica OR the leg pain was simply a "referred pain" coming from the disc [same thing that occurs when someone gets a heart attack and they get pain down their left arm – that left arm pain is a referred pain]. The antibiotics kill off the bacteria which stops the release of propionic acid and tumor necrosis factor alpha, which slowly stops the cause of the pain (irritation of the nerves in the endplate/vertebra).

My Comments:  This theory of bacteria getting into the disc herniation is not new and has been around since the early 2000s.  It was pretty much laid to rest in 2006 when researchers showed that under super sterile conditions (they took the cultures and plated them right in the operating room) no anaerobic bugs were found, which indicated that all these bacteria found in disc herniations were coming from the wound or the air of the operating room and not the general circulation. So I was incredibly surprised to learn about this powerful study which demonstrated that antibiotics significantly reduced the back and leg pain of patients who had failed surgery and/or conservative care. 

There are a few problems, however, which I have e-mailed the author about but have yet to get a response (if I don't soon, I will officially write them up and send them to the editor of the Journal).

#1) since the patients were allowed to take anti-inflammatories and other medication while the study was going on, we need to make sure that there was no difference between the two groups with regard to usage. For example, what if the patient's in antibiotic group were taking more anti-inflammatories/other medications as compared to patients in the placebo group-- that would not be good.
#2) did the blood serum test results demonstrate signs of bacteria? They can test for Lipid S (a marker on anaerobic bacteria), which could have told us if bacteria were present in these patients – that's the biggest problem with this study is that we don't know whether or not bacteria were even present in these patients.
#3) Albert mentioned in the article that not all of the patient's completed the course of antibiotics and placebo, but he did not tell us if the noncompliant patients were the same in number between the groups.

Other than that, this was a very well designed study, and I agree with Albert's recommendations for patients who have failed conservative care/surgery and have a history of symptomatic disc herniation with Modic type I change at the same level, to consider undergoing a 100 day course of antibiotic therapy.

Our Future Research:

We have decided to conduct a study to see if bacteria are really infecting the disc via the circulation--we believe that the bacteria found in HNP material is coming from the surgical incision, not the general circulation. We are randomizing a group of patients undergoing microdiscectomy into either the traditional antiseptic procedure group and a group that will get an extra cleaning of the wound after dissection down to the lamina. Then, as others have done, we will biopsy the disc herniation fragment and compare all the results. We theorized that by super cleaning the wound, the number of bacteria found in the herniated fragment will be virtually nonexistent. hopefully we can get this by our IRB.

Home | Top | Research Corner

Copyright© 2013, Douglas M. Gillard, D.C. – All Rights Reserved