February 21st, 2015

Fusobacterium, Pharyngitis, and the Limits of Limiting Antibiotics

A paper on pharyngitis in young adults, just published in the Annals of Internal Medicine, is creating a controversy in the intersecting worlds of primary care and Infectious Diseases. The first author is Robert Centor, of the famous Centor criteria that assess the likelihood of group A strep. He’s been writing about our need to expand diagnostic considerations in sore throat for several years, starting with this excellent editorial.

The Physician’s First Watch summary of the new paper was spot-on, so I’ll just quote them here (bolding mine):

Some 310 young adults (aged 15–30) presenting with pharyngitis at an Alabama university clinic underwent polymerase chain reaction testing for bacteria in throat swab specimens; 180 asymptomatic students were also tested. Fusobacterium necrophorum was identified in 21% of patients with pharyngitis (and 9% of asymptomatic students), while group A streptococcus was found in just 10% of patients (and 1% of asymptomatic students). Clinical presentations were similar for F. necrophorum and group A strep.

From the perspective of patient management, there are two interpretations circulating about this paper — one that it encourages antibiotic prescribing, the other that it does no such thing.

The controversy is nicely encapsulated in this comment on a listerv (remember those?) for pediatricians, which was shared with me from a very reliable source (she didn’t write the comment):

OMG! So if it is cultured [sic] from the throat, it is the cause of the infection,right? So now everyone who has this in their throat and doesn’t feel well needs antibiotics, right? 

Allow me to take both positions:

Pro Antibiotics: Some really terrible exudative pharyngitis in young adults is group A strep negative. This study shows that fusobacterium is more common than strep in this age group. We know it can cause peritonsillar abscess and, even worse, septic jugular vein thrombophlebitis (Lemierre’s syndrome), both of which are preceded by sore throat — and both of which are more common than acute rheumatic fever. If we treat the really sick teenager and young adults who are group A strep negative with an antibiotic with activity against fusobacterium — penicillin and other beta lactams, please, not azithromycin — not only will these youngsters get better faster, but we can prevent potentially life threatening complications.


Anti Antibiotics: Most pharyngitis is causes by respiratory viruses. There is no way to detect fusobacterium as a cause of pharyngitis in clinical practice, so if most cases get treated “empirically”, this will be massive unnecessary treatment. Detection of the organism by polymerase chain reaction in the study does not prove that fusobacterium is the cause of the pharyngitis, especially since it’s found in a not insignificant proportion of asymptomatic individuals (9%). There is furthermore no proof that treatment of fusobacterium will hasten symptom improvement or, more importantly, prevent Lemierre’s.

The latter position was nicely articulated in an accompanying editorial in the Annals written by my colleague Jeffrey Linder — a primary care physician who has published extensively on this subject and admittedly a much more reliable expert on the topic than I.

But let me risk taking a position slightly different from Jeff and, I’m sure, many of my ID brethren, one that I confess is rooted not so much in data but in experience caring for several young adults with Lemierre’s. Importantly, Jeff and I don’t disagree — it’s more a matter of emphasis.

Remember this — patients with Lemierre’s are often critically ill. They frequently require ICU care, have high spiking fevers with staggeringly high white blood cell counts, and invariably have multiple septic pulmonary emboli with potentially other metastatic sites of infection, including the brain.

It’s a terrifying illness — these are most commonly previously healthy high school and college-age kids, so the stakes are high. No, we don’t know that treatment of severe pharyngitis “caused” by fusobacterium will prevent Lemierre’s, but doesn’t that make biologic sense?

So let’s go with the pediatricians’ common-sense approach to clinical care, and make a decision about antibiotics based on that sixth sense of “is the kid really sick?” If so, go with some penicillin — especially if at the first encounter they didn’t get treated, and then they come back a few days later even worse.

Or, if you prefer, listen to the guru of pharyngitis himself, Dr. Centor, and his interpretation of national guidelines:

We believe that following the American College of Physicians/Centers for Disease Control and Prevention guidelines endorsed by the American Academy of Family Physicians would decrease the risk of Lemierre syndrome in adolescents and young adults. Using these guidelines, physicians can choose to prescribe antibiotics for patients with a pharyngitis score of 3 or 4 (three or four of the following: fever, absence of cough, tender anterior cervical lymph nodes, tonsillar exudate).

Makes sense to me.

11 Responses to “Fusobacterium, Pharyngitis, and the Limits of Limiting Antibiotics”

  1. Loretta S says:

    Is clindamycin a good option to cover both group A strep and f. necrophorum for penicillin-allergic patients? Thanks for this post, Paul.

    • Paul Sax says:

      Yes clindamycin is a reasonable alternative.
      FYI, if they get sick enough for admission, we try to ascertain if they really are penicillin-allergic (most aren’t), then usually give a betalactamase-stable option such as amp-sulbactam or piperacillin-tazobactam.

  2. Thanks greatly for your interpretation of our study. You have emphasized the major point. While uncommon (we estimate 1 in 70,000 adolescents/young adults each year), the Lemierre syndrome is devastating. If we believe in preventive medicine, then this clearly is a disease that we should work hard to prevent.

    We also must continuously remind all physicians that we are focused on this age group and NOT PRE-ADOLESCENTS. We are talking about treating approximately 30-40% of adolescent/young adult pharyngitis with antibiotics that have not lost their efficacy despite over 50 years of use.

  3. Sanford Kimmel says:

    Speaking from the standpoint of a college student health physician, I not infrequently see young adults presenting with exudative pharyngitis, often with fever. If the rapid strept is negative, my next step is to order a monospot which may be positive in our population. If this is negative, I will often then place the patient on oral penicillin V. My point is that infectious mononucleosis also needs to be considered in the differential diagnosis.

    • My point is that infectious mononucleosis also needs to be considered in the differential diagnosis.

      You make a very important point. Some authors have suggested that Infectious Mononucleosis patients also have increased risk of Fusobacterium necrophorum pharyngitis and thus the Lemierre Syndrome. Finding infectious mono does not exclude Fusobacterium necrophorum.

      Infectious mono is a very important part of the differential, as is acute HIV infection. Adolescent/young adult pharyngitis is not a simple problem. Thanks for your thoughtful comment.

  4. […] Paul Sax, in a current blog post, explains our position […]

  5. Richard Davey says:

    The authors make a compelling case for better understanding of the role of F. necrophorum in serious pharyngitis. However, I am surprised that the editorial accompanying the primary article did not challenge the authors’ assertion that performing an RCT of tx vs no tx in patients with F. necrophorum carriage/infection would be “unethical”. In short, we have an association that currently goes largely undiagnosed and untreated due to lack of a commercial detection assay, in the setting where the preventive value of treating patients with this isolate remains totally unproven, yet somehow this unfortunate state of “ignorance is bliss” is preferable to actually studying in an RCT whether active tx decreases the rate of complications from this carriage/infection such as the uncommon Lemierre’s Syndrome? To the contrary, this is actually the setting where an RCT is needed.

    • How many patients would one need to enroll to prove that antibiotics prevent the Lemierre syndreom? In my 2010 Annals of Internal Medicine article, I estimated that 1 in 400 patients with F. necrophorum pharyngitis would develop the Lemierre syndrome. Thus, you can extrapolate how many patients would need randomization to prove efficacy or lack of efficacy.

      Would you recommend your adolescent child with a severe sore throat enter such a randomized controlled trial? I would not. We are really not talking about ignorance. We are really arguing the value of circumstantial evidence.

      Perhaps we could follow several thousand patients testing positive for F. necro (after we have a point-of-care test) and see if the develop complications. But having evidence of infection and withholding antibiotics would put the investigators at legal risk and the patients at unnecessary medical risk.

      I stand by my original assertion.

      • Richard Davey says:

        I stand by my original assertion.

        As do I. While your focus appears to be on the uncommon complication of Lemierre’s syndrome, the fact is that a properly designed RCT could address whether identification and specific antibiotic treatment of this previously under-diagnosed pathogen actually makes sense in terms of affecting a significant reduction in overall morbidity of severe pharyngitis (e.g. duration of symptoms, days of school lost, secondary complications, etc.) compared to watchful waiting (i.e. the current approach in strep-negative patients). A change in guidelines demands no less, and studies of several thousand patients are hardly unusual. And yes, while such arguments are rarely helpful, I would enroll my child in such a trial since right now we have no idea whatsoever whether prompt antibiotic therapy offsets the incidence you estimate from your prior work. You would not be putting such patients at any higher risk than the current approach of doing nothing simply because this particular pathogen is not being diagnosed by current techniques.

  6. Louis Menachof M.D. says:

    infectious mono involves the posterior cervical chain of nodes in addition to others. Group A Beta strep ( and probably F. Necrophorum) involve only the superior nodes of the anterior cervical chain (commonly referred to as the “tonsillar nodes). This should help distinguish between the causes of adenopathy.

  7. Michelle M Condon MD says:

    The video was removed from the site…bummer

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.