Learning Objectives: After reading this article, the physician should be able to:
1. Discuss the physiology of malignant otitis externa in the diabetic patient.
2. Identify the clinical presentation of malignant otitis externa in the diabetic patient.
3. Summarize the treatment of malignant otitis externa in the diabetic patient.
The inexorable diabetic process results in anatomic and biochemical changes in the foot of diabetics that allows few to escape such foot problems as ulcers, infection, and sometimes amputation. A combination of autonomic, motor, and sensory neuropathy, coupled with the underlying immunocompromised state made worse by vascular insufficiency, creates the perfect storm for such ulcers to develop and to heal poorly, slowly, or not at all.
Emergency physicians can do little to treat these chronic problems in the ED, but we can identify them, and we certainly must respect them. Careful attention to detail, including specific follow-up and expeditious contact with a consultant, are paramount to keep the diabetic out of the operating room and the emergency physician out of the courtroom. A liberal admission policy is suggested for diabetics who may otherwise have an outpatient problem but lack the money for antibiotics, a car for trips to the wound care center, a concerned family and a clued-in primary physician, or the eyesight necessary to assess their own healing.
This month's column discusses an unusual infection that favors diabetics — malignant otitis externa, also known as invasive otitis media. Most clinicians can recite chapter and verse about malignant otitis in diabetics, but few have seen it, and the topic is on every board examination. Similar to a nascent foot infection that turns a minor problem into an amputation, otitis externa can likewise run amuck, and do so very quickly. The emergency physician has to realize that while otitis externa is a minor annoyance to most individuals, it can be a disastrous infection for the diabetic.
Malignant Externa Otitis: A Severe Form of Otitis in Diabetic Patients Zaky D, et al Am J Med 1976;61:298
This commonly quoted article discusses the pathogenesis and treatment of malignant otitis externa in diabetics. It was one of the first articles to focus on a problem that now has widespread attention in the literature. Malignant otitis externa is a severe form of an otherwise benign infection that is occasionally encountered in elderly diabetic patients. It is caused by Pseudomonas aeruginosa. If this disease is unrecognized, it usually progresses to serious complications, including deep infection of the cartilage, bone, nerves, and soft tissue surrounding the ear. It may eventually culminate in osteomyelitis of the skull. Cranial nerve palsy can occur, and this is a harbinger of a complicated course. A fatal outcome is almost certain if this disease is allowed to progress on its own. The disease is described in two case reports that were seen at the authors' institution.
The first case was a 75-year-old man who first presented with a three-week history of severe pain and discharge from the right ear canal. He was a diabetic of only six years duration, and was well controlled by oral therapy. He had intermittent mild problems with otitis externa of both ears, but they were refractory to a variety of standard outpatient treatments. Prior to this admission, he had a known Pseudomonas, Enterocossus, and Candida albicans externa otitis that had been treated with silver nitrate drops and oral ampicillin. Because of the poor response of the chronic infection, he was admitted for mastoidectomy. He required two additional admissions, but the infection still was not totally cured. A right facial palsy developed weeks after initial therapy, and his ear infection was never completely cured prior to his death from a CVA.
Although most EPs will never see a case of malignant externa otitis, it's a serious disease that has to be treated with respect
The second case was a 57-year-old woman who complained of severe pain and swelling of the right ear. She also was a diabetic whose blood glucose was well controlled on oral therapy. She had previous disastrous problems with her diabetes, including blindness, leg amputation, and deafness. On this presentation, she exhibited a polypoid swelling of the externa canal, a relatively common finding in malignant otitis. The pain continued for a number of months despite empiric therapy for a presumed Pseudomonas infection. Her condition also was complicated by facial nerve involvement and recurrent infections despite surgery, and the use of the most potent intravenous antibiotics of the day. Her case was thought to be unusual because Pseudomonas was never cultured at any time from the externa ear discharge, but the organism was found at surgery when an abscess cavity was discovered.
A number of clinical features are associated with malignant externa otitis. It is a rare entity, but is most common in elderly diabetics who have reasonable blood sugar control with oral agents. The hallmarks are severe pain and discharge from the ear. Cultures invariably demonstrate Pseudomonas. When this paper was published nearly 30 years ago, only about 60 cases had been reported in the English literature.
This condition in the ear is equivalent to diabetic gangrene of the limbs in severity and outcome. The deep recesses of the ear and the anatomy around the skull make Pseudomonas otitis externa difficult to cure, and allows it to spread to many nearby structures. Curiously, facial nerve palsy is described in about half of the persistent cases, and it's usually the first sign of neurological involvement. A seventh cranial nerve palsy can be secondary to infection at the stylomastoid foramen or secondary to osteomyelitis of the bony canal itself. It is a poor prognostic sign, and is often associated with mortality. The cases in the literature demonstrate a 50 percent mortality rate once cranial nerves are affected.
Cultures generally demonstrate pure Pseudomonas aeruginosa, but occasionally mixed cultures with Staphylococcus, Enterococcus, and Candida are noted. Rarely Pseudomonas will not be cultured from the canal, but will be the culprit of a deep infection found at surgery. Although Pseudomonas is commonly cultured in otitis externa in healthy individuals where it does not seem to be a major problem, diabetics have a predisposition to this organism that results in a relentless progression. Unlike swimmer's ear, Pseudomonas in the ear canal in a diabetic does not respond well to topical therapy.
The authors highlight the characteristic granulation tissue in the ear canal that is commonly associated with this unusual ear infection. Importantly, patients are generally afebrile, and the white blood count is only slightly elevated or normal. Most patients require hospitalization and systemic antibiotics. The poor vascularity of the involved areas hinders penetration of antibiotics. Meticulous surgical debridement of all infected tissue should be performed early in the course of this disease.
Comment: Otitis externa is a severely painful process in most individuals. It is commonly known as swimmer's ear, and it's omnipresent in the ED. Otitis externa is a relatively straightforward diagnosis, and standard treatment, often only ear drops, is usually quite successful within a few days. Occasionally a few days of oral antibiotics hasten improvement. It's an easy diagnosis to make. Pressure on the tragus or manipulation of the ear pinna causes the patient to jump. The otoscope exam is painful, and demonstrates redness, swelling, and discharge from ear canal. The clinician can readily make the diagnosis at the bedside. Generally the debris from the ear canal of normal individuals is not cultured, but when it is, Pseudomonas is not an uncommon finding. Staphylococcus is another commonly cultured organism. The Pseudomonas organism is more of a curiosity than a great clinical concern in most healthy individuals, but in a diabetic, this bug is a recipe for disaster. Patients with HIV also are fair game for invasive otitis externa.
In reality, malignant otitis externa is quite uncommon. We all know about it, we all know it's caused by Pseudomonas, and we all know it's a serious disease that should be treated with respect, but many clinicians will not see a single case in an entire career. The propensity of the process to involve cranial nerves and the base of the skull attests to the need to highly respect this infection in diabetics. One proposed reason for why Pseudomonas is a minor pathogen in most ear canals yet a prognosticator of disaster in the diabetic is that the micorangiopathy of diabetes results in hypoperfusion of the skin of the ear canal. It sounds like the same underlying problems with diabetic feet that predisposes them to foot ulcers and infections.
Diabetics who have a mild case of otitis externa of short duration can be treated similarly to other patients if they are not toxic or particularly ill. Corticosporin is appropriate for most mild cases, but I like ofloxacin drops for all patients with a significant otitis. Ofloxacin drops are effective and safe, even for a perforated tympanic membrane, but these drops are quite expensive (Drug Safety 2003;26:405). Make sure the prescription plan covers them. One can try a short course of antibiotic drops alone, but I frequently add oral antibiotics to most patients who have significant otitis because it is actually a cellulitis. Most cellulitis in other areas of the body is not readily cured with topical agents, so it seems to make good sense to add systemic therapy. Keflex or dicloxacillin are recommended therapies. The cellulitis in normal individuals is usually caused by Staphylococcus aureus, and because it's only a minor irritation or infection, most patients experience a rapid effervescence of both pain and swelling. For the diabetic, ciprofloxin is the oral antibiotic of choice for outpatient otitis externa.
Most cases of invasive otitis externa make themselves known with gusto; the pain makes even the most stoic diabetic cry. Some cases are more chronic and indolent, however, and seem like a minor problem that just hangs on and smolders for weeks. These are probably the most serious cases because they don't get intense medical attention until the infection becomes deep-seated.
I see no sense in ordering a CBC in any case of otitis externa, and don't expect this test to be predictive or prognostic of serious malignant otitis in the diabetic. Allowing the presence or absence of fever, leukocytosis, or culture results to guide your therapy in diabetics with otitis externa demonstrates a lack of knowledge of the physiology of this process. Also, don't expect to see ketoacidosis or a blood sugar of 800 with this ear infection. Because the otitis itself occasionally may be unimpressive, any pain in the side of the head should be evaluated for malignant otitis externa. A Bell's palsy in the diabetic calls for a good ear exam, but the ninth, 10th, and 11th cranial nerves also have been involved.
Diabetics with otitis externa who are initially treated as outpatients should be rechecked in 48 to 72 hours with a definite planned appointment. “See your doctor for follow-up” is appropriate for the teenage swimmer with a sore ear canal, but not for the diabetic with otitis externa. If pain is severe and the canal is swollen shut or if there is significant cellulitis of the externa structures, it's prudent and reasonable to admit these patients at the first visit and be done with it. Empiric intravenous antibiotic effective against Pseudomonas can be started in the ED. In my experience, most patients with a rip-roaring otitis externa require narcotics for pain control, and it's difficult to send them home when they are crying in agony. An ENT and ID consultation are expected as part of the hospitalization, and diabetics usually end up with an MRI or other tests to rule out mastoiditis, abscess formation, or osteomyelitis. Should you see cranial nerve involvement, such as a seventh nerve palsy, your patient is in for trouble.
Skull-Based Osteomyelitis: Malignant Externa Otitis Slattery W, Brackmann D Otolaryngologic Clin NA 1996;29(5):795
This is an extensive review of malignant externa otitis that emphasizes the potential for this skin infection to progress to osteomyelitis of the skull. Other complications reviewed include meningitis, cavernous sinus thrombosis, and cranial nerve palsies. The requirement for intensive long-term antibiotic therapy and periodic radiographic evaluations to monitor this process are discussed. Because of the potential for the development of resistance to antibiotic therapy, Pseudomonas osteomyelitis in the diabetic still has a high mortality rate. The authors note that some other processes mimic skull-based osteomyelitis, including carcinoma, tuberculosis, Paget's disease, and fibrous dysphagia. Occasionally, osteomyelitis is the prominent first manifestation, and the initial ear infection is not appreciated.
Granulation tissue in the externa ear canal is a common finding in malignant otitis, and is correlated with osteomyelitis of the skull. The authors emphasize that granulation tissue is the most important indicator on physical examination of malignant otitis.
Antibiotic therapy is the mainstay treatment of Pseudomonas infections. Anti-Pseudomonas penicillin, such as carbenicillin, are less useful because of inactivation by beta-lactamase. Gentamycin was a prior mainstay in the treatment of Pseudomonas infections, as was tobramycin. Both drugs can cause ototoxicity and nephrotoxicity, and their use, especially as single agents, has been greatly curtailed. Aminoglycosides as monotherapy for severe Pseudomonas infections anywhere in the body is not considered appropriate.
The two drugs commonly used for Pseudomonas infections are the third-generation beta-lactamase cephalosporin, ceftazidime (Fortaz), and the quinolone, ciprofloxacin (Cipro). Imepenem and meropenem also are reasonable empiric monotherapy. Resistance of Pseudomonas to quinolones and all other drugs is increasingly problematic. If possible, a culture of malignant otitis should be obtained to confirm Pseudomonas and to assess drug sensitivity. Some authors advise no internal ear drops and sequential cultures to assess sensitivities during prolonged treatment.
The value of hyperbaric oxygen remains unclear, but it is commonly used in severe infections if it is an available modality.
Comment: These authors and others comment on the curious finding of granulation tissue in the canals of diabetics with malignant otitis. It's not uncommon to find some sort of debris in the canal of anyone with otitis externa, but usually it is minimally problematic. I have not personally seen this granulation tissue, but it's mentioned in all of the reviews. Usually the canal is so swollen during the first few days of any serious otitis externa, one cannot even see into the canal. Although it would be unusual, the presentation of malignant otitis as a cranial nerve palsy or skull osteomyelitis without an impressive externa otitis has been noted before.
Treating presumed Pseudomonas infections used to be simple: Gentamycin killed everything and was ideal as long as the kidneys and ears were not damaged. Now things are more complicated, and no hospital antibiotic sensitivity chart from the lab has Pseudomonas 100% sensitive to any single antibiotic. Lab sensitivities and clinical cures do not always equate, however. Culturing an externa otitis is not routine, but because of the potential for antibiotic resistance, cultures have usefulness in this scenario. Initial treatment of otitis externa in the diabetic would be empiric, but certainly should cover Pseudomonas.
It's a good idea to assume that gentamycin and tobramycin alone will not cure a serious Pseudomonas infection. Combination therapy is required. A good empiric choice in the ED would be imepenem, meropenem, cefepime, Zosyn, ceftazidime, or a quinolone. Cipro seems to be the preferred quinolone, with lower MIC against Pseudomonas than Tequin or Levaquin. In the very septic patient, I would pick two from this list or add an aminoglycoside until the infectious disease consultant sees the patient. The ID consultant in our hospital likes cefepime (Maxipime) alone as the initial choice for malignant otitis externa in an adult.
It is probably impossible to perform any sort of meaningful debridement of the ear canal in the diabetic or any other patient with otitis externa, and I would not even try to initiate it in the ED. I think the textbook admonition of debriding the canal to assure treatment success with otitis externa is a myth. Most patients do just fine with drops and antibiotics without the ear curette causing torture. Diabetics, on the other hand, may require general anesthesia for an adequate debridement and evaluation of the deeper structures.
It is a common scenario that otitis externa follows irrigation of ear canals for impacted cerumen. When I irrigate wax out of anyone's canal, I usually give them a few days of eardrops to prevent otitis externa. I would prescribe ofloxacin eardrops after irrigating wax from a diabetic canal, and vosol (2% acetic acid) after irrigating the ear in others. Vosol is effective and inexpensive, but it's an eardrop not known to most emergency physicians. One reference notes that malignant otitis was preceded by irrigation of an ear canal with tap water in eight of 13 diabetics. A substantial number of cases are likely iatrogenic or self-induced by the patient trying to clean the ears (Ann Otol Rhinol Larynagol 1990; 99(part 1):117).
HBO might help patients with invasive otitis. Data are hard to obtain, but some have reported anecdotal success with multiple treatments (Arch Otolargynol Head Neck Surg 1992;118:89.