October 21st, 2019

Amoxicillin for Chronic Low Back Pain? Are You Kidding Me? In Defense of a Controversial Clinical Trial

The BMJ just published a randomized trial comparing amoxicillin to placebo for people with chronic low back pain.

I kid you not.

The appearance of this trial elicited all kinds of snark from the medical community. Here, check this out, along with the responses:

Dr. Juurlink, definite points for that <checks notes> stage direction! Love it.

But allow me to defend the people who did the study, and go even further — this is exactly the sort of practical, hypothesis-testing trial I wish we’d see more often.

Consider the problem — chronic low back pain. The bane of Western Civilization, it occurs in a quarter of the adult population. The misery from this condition results in millions of annual office visits a year, countless days out of work, and heavy economic losses.

Usually we don’t know the cause. And for severe sufferers, our medical treatments and surgeries offer inconsistent benefits.

Along comes the idea that a subset of back-pain patients — those with certain inflammatory changes on imaging, called “Modic” after the person who described them — might have a low-grade infection as the cause of this inflammation.

The theory is that a degenerating disc provides a suitable spot for this infection to settle, presumably after transient bacteremia. It would have to be a very indolent, slow-growing infection, as people with chronic low back pain don’t have fevers or other symptoms of acute infection, and also lack laboratory evidence of infection or inflammation.

Hi there. Nice seeing you again. Aren’t I a cutie-bacterium?

Based on animal and human data, the leading candidate for this kind of infection is none other than our old friend Cutibacterium acnes, shown here again in a very good mood.

(Maybe it’s happy because, as I’ve written before, Cutibacterium acnes used to be Proprionobacterium acnes. The new name is a zillion times better, especially if it’s pronounced like a cute little puppy, and not like a paper cut, or like an old coot.)

We ID doctors are quite familiar with C. acnes as a relatively common cause of prosthetic joint infections, especially of the shoulder; dermatologists know it as one of the primary bacteria involved in <checks notes> acne.

C. acnes also sometimes pops up in blood cultures, usually as a contaminant — but maybe those aren’t contaminants after all!

(Cue dramatic music here.)

The idea that C. acnes could contribute to chronic low back pain has been supported by the occasional isolation of the organism during spine surgery, and animal models showing the bug could induce these Modic changes in rabbits. This information led to a controversial randomized clinical trial comparing amoxicillin-clavulanate to placebo in adults with chronic low back pain, showing significant improvement in the treatment arm.

A subsequent systematic review concluded:

… further work is needed to determine whether these organisms are a result of contamination or represent low grade infection of the spine which contributes to chronic low back pain.

All of which brings us back to the recent study, which enrolled 180 people with chronic low back pain and Modic changes (of two types) on imaging. They were randomized to receive oral treatment with either 750 mg amoxicillin or placebo three times daily for three months. The primary outcome was a validated disability score a year later. They set a difference of 4 points on the scale as being clinically meaningful.

The results showed that the amoxicillin group had significantly lower disability scores than the placebo group, but the difference did not meet the threshold for being clinically important (it was only 1.6 points). Plus, nearly twice as many in the amoxicillin arm experienced a drug-related adverse event.

I certainly agree with the authors’ conclusions that the “results do not support the use of antibiotic treatment for chronic low back pain”, especially if you consider the added potential problems of encouraging antibiotic resistance and alteration of the human microbiome.

But kudos to them for doing the research — even negative studies are important. It could have been an H. pylori and peptic ulcer disease, but was C. pneumoniae and CAD instead.

But just imagine if it worked!


15 Responses to “Amoxicillin for Chronic Low Back Pain? Are You Kidding Me? In Defense of a Controversial Clinical Trial”

  1. Eugene Fung, MD, FRCPC,FACR says:

    I applause Dr. Sax on bringing this topic,
    As for chronic pain the practitioners have been preoccupied with disc protrusion or herniation, foraminal root encroachment, but surgical treatments are often not solving them except the ones with good indication.
    Modic change as well as the discitis have been mostly neglected. The previous usage of TNF inhibitor and the current usage of Amoxicillin are all thinking outside of the box.

    That might also remind us how busy in the past that gastric ulcers were treated with gastrectomy and V & P.

  2. Andrew Tsoi MD says:

    If one were NOT to automatically think of Amox as an “antibiotic” but a good “anti-inflammatory”, one’d see the logic of the data conclusion. likewise Doxy and Trimethoprim

  3. Clancy Hughes, m.d. says:

    What about salmonella and variations of spinal, peripheral spinal and Reiter’s arthritis? Does amoxicillin have a role, and are stool cultures helpful?

  4. Jayant Thorat says:

    Along with MRI contrast a leukocyte scan should be prescribed in thise with new modic changes and back pain before we run more trials.
    If infection has subsided then amox may not be beneficial.Pain though may persist due to degenrative changes causing disc ht reduction or annular tears

  5. Carlos E. Goicochea says:

    Brucellosis was ruled out?

  6. Chris Stoner says:

    Another humbling experience. My first thought, too, was what were they thinking. The explanation makes perfect sense and again reminds me of my knowledge deficits.

  7. GHassan Al Awar says:

    Hello Paul,
    As I went through the text, brucellosis came pressing to my mind.
    In the Middle East we are quite familiar with brucellosis as a cause of a febrile illness, chronic or recurrent, with musculoskeletal symptoms esp backache, hip & thigh pains. Children may be diagnosed after limping! Brucella spp may respond to many (or any) Abcs on condition that those are given in combination and for long enough (usually 6 wks).
    In that case a response to amoxicillin is not far fetched esp if given long term!
    So, I agree with the last response above of Carlos: “Was brucellosis ruled out?”

  8. GHassan Al Awar says:

    One more question Paul, what do you think of Dilaudicillin??

  9. Steve Kalish says:

    Just a reminder to the readers who acquiesce to performing junk science: Vitamin C does not work for sepsis.

  10. Sarantos Soumakis, MD says:

    Did not read the actual study, but can someone explain why Amoxicillin was chosen as treatment for the presumed cause of chronic low back pain as presented in the study?
    Amoxicillin is certainly not a first line antibiotic for Acne.

  11. Jason Helm says:

    Thank you, this was very informative.
    Being a patient of lower back pain I have been to mostly all the doctors and their medicines seemed to help me temporarily. Currently my doctor is of pain management doctors in rockland county ny
    and he has recommended me thermotherapy. Which has been quite helpful for me. The icing might help in injuries but I believe it would cause more pain if applied for joint pain.

    • Les Peterson, DC, RT, ARRT says:

      Icing often increases muscle guarding and pain. Misapplication of heat can also exacerbate splinting and increase pain. The selection of ice or heat should always be a decision the provider makes on a patient by patient basis.

  12. Les Peterson, DC, RT, ARRT says:

    Since most of the low back pain seen in primary medical settings is mechanical in origin, and Chiropractic manipulative therapy provides benefit in a cost effective manner, perhaps a referral to a chiropractor would help empty these troublesome patients from waiting rooms where questions of etiology can run to subtle infection before mechanical dysfunction. Indeed, many work compensation stakeholders have seen that an attending DC treating simple mechanical back pain rather than an attending MD reduced risk of total disability by a significant margin.
    What ever happened to “hearing the sounds of hooves” should evoke thoughts of horses before zebras?

  13. Moses says:

    In order to solve the problem of sedentary low back pain, you must buy a comfortable chair to work in the office.

  14. Nancy Ramsdell says:

    I have chronic low back pain. (It is moving up my spine) I know Penicillin VK will help (maybe cure) my chronic pain. Pen VK also crosses the blood brain barrier. It works faster than amoxicillin. I can’t find anyone who will prescribe this for me. If you are willing to help me, e-mail me. I have battled this for 8 years. I have been battling this for 8 years.

HIV Information: Author Paul Sax, M.D.

Paul E. Sax, MD

Contributing Editor

NEJM Journal Watch
Infectious Diseases

Biography | Disclosures | Summaries

Learn more about HIV and ID Observations.