Chronic suppurative otitis media: A clinical overview.

There is no widely agreed-upon definition of chronic suppurative otitis media (CSOM). I consider CSOM to be purulent otorrhea that persists for more than 6 weeks despite treatment. This definition is not one that you most often see in the literature or one that is used in clinical trials because, technically speaking, a chronic tympanic membrane perforation that is clean and dry is often referred to as chronic otitis media. By my definition, mastoiditis is invariably a part of the pathologic process, but cholesteatoma is not.

Pathogenesis

CSOM can occur in two ways. Some patients develop CSOM as a result of an earlier episode of otitis media that did not resolve completely, often because initial treatment was incomplete or inadequate. Other patients develop CSOM as a result of a pre-existing tympanic membrane perforation (which could occur spontaneously or following placement of a tympanostomy tube). A perforation or tympanostomy tube eliminates the "middle ear cushion" by allowing air to escape from the middle ear space, which allows nasopharyngeal secretions to reflux into the middle ear, especially in smaller children.

The onset of CSOM is characterized initially by increased vascularity of the mucosa and submucosa. As CSOM persists, the proportion of chronic inflammatory cells increases. This leads to osteitis and mucosal edema with ulceration. Two important pathologic events follow mucosal ulceration: (1) capillary proliferation, which results in the formation of granulation tissue and polyps, and (2) a rarefying osteitis, which ultimately produces new bone formation and fibrosis. Osteitis is present in virtually 100% of CSOM patients, a finding that distinguishes CSOM from more transient pathologic alterations of the middle ear cleft.

The issue of anaerobic organisms is still controversial. They are usually present, but whether or not you will find them depends in large part on how hard you look for them. No one is certain as to the pathophysiologic significance of the anaerobic component.

Management

At the University of Texas, our strategy for managing CSOM is to start with an aural toilet regimen, topical antibiotics, and, if necessary, control of granulation tissue. If this is unsuccessful, we obtain a culture because this is the point where we might consider systemic therapy. It is also the point where we extend our evaluation to include an immune work-up for immunoglobulin subclass deficiencies and a search for accompanying sinusitis, reflux, or allergic rhinitis.

Given the microbiology of CSOM, the options for systemic treatment are usually limited to intravenous aminoglycosides, extended penicillins, oral or intravenous cephalosporins that cover gram-negative organisms, and oral fluoroquinolones.

Critical to the management of CSOM is the control of granulation tissue. In fact, I believe that granulation control might be the principal benefit of surgery. All of us who have operated on a patient with granulation tissue know it is much harder to deal with than cholesteatoma.

The control of granulation tissue requires the administration of an appropriate anti-infective, either a quinolone or an aminoglycoside. Most physicians, myself included, believe that steroids are helpful despite the fact that there are no hard data to support their use. Nevertheless, this prejudice is widespread and deeply ingrained, and I am a firm believer in using steroids to control granulation tissue.

One can also control granulation tissue with various types of cautery. In the operating room, we use electrocautery. In the office, we are more likely to use chemical cautery, usually with silver nitrate. But one must be careful with cautery because there have been reports of patients who experienced facial nerve paralysis following silver nitrate cautery, some of whom did not recover.

Polyps (a form of granulation tissue) can be removed in the office with sharp instrumentation. Excising polyps is probably a better technique than simply avulsing them, because polyps can be firmly attached to the ossicles or chorda tympani.

Treatment difficulties

Drug delivery. Why is CSOM so difficult to treat? I believe that part of the reason is poor delivery of medication. Even systemic delivery is not particularly effective in reaching the middle ear and mastoid mucosa. There is an ongoing study at our institution to look into this. We are dosing patients with various quinolones for 5 days and then performing a biopsy to measure drug levels in the middle ear and mastoid mucosa.

Topical therapy, of course, has its own limitations with respect to penetrating into the middle ear space and mastoid. However, there are some interesting studies from Japan suggesting that we really can achieve very good delivery of topical medication through a perforation. (1)

Antibiotic resistance. Development of antibiotic resistance has not been shown to be much of a factor in persistent CSOM. I doubt that it is the cause of treatment failure in more than 1 or 2% of patients.

Bacterial survival mechanisms

Nonplanktonic growth modes. We do not know whether nonplanktonic modes of growth are important in CSOM. Bacteria have many ways of dealing with starvation. One of them is their ability to execute rapid multiple divisions that produce ultramicrobacteria. Ultramicrobacteria are very small organisms that contain a genome and not much else. They exist in almost a spore-like state; in fact, we can regard these organisms as a type of spore formation. When conventional bacteria divide this way, the number of ultramicrobacteria can exceed the number of original bacteria by as much 15-fold.

There is a specific gene--the stringent response gene--that is involved in controlling the division process. This gene apparently responds to an absence of nutrients and other favorable growth conditions in the environment. The starved cells might have meaningfully different phenotypic expressions than do their well-fed counterparts. Because of alterations in their surface structure, starved cells can begin to interact differently with other cells in the environment and can develop the ability to attach themselves to different cells. These surface changes might also affect how antibiotics bind to them, for better or for worse therapeutically.

Biofilms. The development of a biofilm is another mechanism in which starved cells attempt to survive. (2-6) A biofilm is a sessile form of bacterial growth that attaches to surfaces and has functional heterogeneity--that is, bacteria cells can have different phenotypic expressions in different parts of a biofilm. Channels within a biofilm allow bacteria in one part of the matrix to communicate with bacteria in other parts. Some of these messengers have been defined. A biofilm can contain a single species of bacteria or multiple species. The classic example of a biofilm is dental plaque, which contains multiple species.

There are obviously significant differences between planktonic bacteria and the sessile bacteria in biofilm. The planktonic forms have the ability to rapidly multiply and spread, something that organisms in a biofilm cannot do. On the other hand, planktonic forms expose themselves to noxious environmental stimuli, macrophages, bacteriophages, biocides, antibodies, and antibiotics. Sessile forms are basically resistant to these types of risks.

A biofilm is capable of altering its environment in several ways. For example, it can form a glycocalyx roof over the colony that protects it from noxious influences. A glycocalyx is manufactured from a tangled knot of exopolymers, usually polysaccharides and often something as simple as chained sucrose or fructose. Once the glycocalyx is formed, the biofilm concentrates enzymes that erode mucosa and bone. Once these enzymes are concentrated, they can penetrate the mucosa and bone. One can wonder to what extent this is the ultimate cause of some of the mucosal alteration that initiates the process of CSOM development.

In their starvation-mode state, biofilm bacteria are dormant. They are doing nothing more than trying to stay alive. They remain in this dormant state until a type of hormonal mediator signals that environmental conditions are right. Once this occurs, parts of the biofilm can spread by detaching themselves from the colony, dispersing, and attaching themselves elsewhere. They disperse by reverting to their planktonic form, thereby becoming once again vulnerable to conventional antibiotic therapy.

Microbiologists base their antibiotic sensitivity studies on laboratory analyses of naked mutants (i.e., planktonic forms) that are probably not typical of the bacteria that exist in the patient. Nevertheless, most of the clinical information we now have on these cells is derived from microbiologic study rather than clinical study. Two of the most studied organisms in biofilms are Pseudomonas aeruginosa and Staphylococcus aureus--the two primary culprits in CSOM.

Organisms in biofilms are significantly less sensitive to antibiotics than are the same organisms in their planktonic form. P aeruginosa and S aureus might be 20 to 100 times less sensitive to tobramycin in a biofilm state than in a planktonic state. There are several possible reasons for this. It could be that anabolically inactive organisms are innately less sensitive. Or it could be that antibiotics diffuse more slowly through the protective glycocalyx; this could be a significant factor if the enzymes in the glycocalyx are destroying the antibiotic faster than the antibiotic can penetrate the glycocalyx shield.

It has been postulated that the protective property of a biofilm glycocalyx is evident in the difficulty that patients with cystic fibrosis have in controlling pseudomonal pulmonary infections. Oddly enough, a high antibody titer to P aeruginosa in a cystic fibrosis patient is a negative prognostic sign. These patients are producing antibodies properly, but the antibodies cannot get to the causative organism, possibly because the glycocalyx is preventing it. Because any planktonic forms of bacteria that these patients have are rapidly killed by the antibiotic, treatment can lead to a symptomatic improvement even though the infection has not been eradicated.

Post reported that biofilms have been isolated on tympanostomy tubes taken from children. (2) Berry et al identified biofilms in an animal model and in vitro. (3) Berry et al made two other important observations:

* A biofilm can be prevented by an ionic bombardment of the tube, presumably because this reverses the negative electrical charge in the glycocalyx.

* The nature of a tympanostomy tube's surface might be a determining factor in whether a biofilm forms.

Tubes with certain coatings are more resistant to biofilms than are tubes with other coatings. For example, a silver nitrate coating does nothing to prevent a biofilm. The characteristics of a tube's surface are related to the initial attachment phase of the molecules that initiate biofilm development, and therefore they have some clinical significance. I do not believe it is a coincidence that post-tympanostomy tube otorrhea might play a part in the initiating event for many children with CSOM.

Biofilms in CSOM

How are biofilms involved in CSOM? It could be that exposed osteitic bone is a good point for biofilm attachment. It could be that biofilms cause mucosal ulceration and bone erosion by attaching themselves to an altered cell membrane that has been infected by a virus. It could be that biofilms attach themselves to foreign bodies in the ear. Foreign bodies in the ear are not very common preoperatively, but they are postoperatively--tympanostomy tubes being an obvious example. Other foreign bodies might include metal fragments of a suction tip that fall into the mastoid cavity during surgery. Finally, it could be that there are important cell-cell interactions, especially if those cells have been altered by viral infection.

What can we do about biofilms in patients with CSOM? There are several strategies that might someday be used to either interfere with biofilm formation or eradicate it once it has been established, including receptor blockade, the introduction of antipolymerases, and interference with biofilm messengers (table).

With regard to receptor blockade, it has been shown that treating mouse peritoneum with glycocalyx polysaccharides will prevent Bacteroides fragilis from forming a biofilm. The polysaccharide will occupy and block the B fragilis receptor sites so that the bacteria cannot attach to them.

The introduction of antipolymerases might break up the protective glycocalyx roof. This could be accomplished by competitive enzyme inhibition that presumably would not affect normal body tissues.

Another option might be to identify the biofilm messengers and use them to send a message to the biofilm that it should revert to its planktonic form, which could then be easily eradicated with conventional antibiotics.

Comments

Dr. Croxson: Is there a place for dry-powder insufflation in the management of CSOM?

Dr. Roland: I am an advocate of powders, and I use them frequently. Powders are now virtually the mainstay of my treatment for patients with a chronic mastoid cavity otorrhea and for those patients with CSOM who have large perforations. Although there are no published scientific data on this issue, there are many physicians like myself whose experience tells us that powders are significantly superior to liquid antibiotic agents. This might not seem like particularly compelling evidence for most physicians, but there is much to be said for everyday clinical experience.

Table

Potential interventions for biofilm-mediated infections

Attachment-site receptor blockade
Disruption of synthesis
 (antipolymerases)
Synthetic biofilm chemical
 messages (i.e., "antiquorum-
 sensing chemicals")
Mechanical disruption (surgery)
Detergents
Penetrance enhancers
Pulse antibiotic therapy
Mechanical surface alternatives
Biocide-containing surfaces

References

(1.) Ohyama M, Furuta S, Ueno K, et al. Ofloxacin otic solution in patients with otitis media: An analysis of drug concentrations. Arch Otolaryngol Head Neck Surg 1999;125:337-40.

(2.) Post C. Direct evidence of bacterial biofilms in otitis media. Laryngoscope 2001;111:2083-94.

(3.) Berry JA, Biedlingmaier JF, Whelan PJ. In vitro resistance to bacterial bioflim formation on coated fluoroplastic tympanostomy tubes. Otolaryngol Head Neck Surg 2000;123:246-51.

(4.) Costerton JW, Geesey GG, Cheng KJ. How bacteria stick. Sci Am 1978;238:86-95.

(5.) Stewart PS, Costerton JW. Antibiotic resistance of bacteria in biofilms. Lancet 2001;358:135-8.

(6.) Lewis K. Riddle of biofilm resistance. Antimicrob Agents Chemother 2001;45:999-1007.

Peter S. Roland, MD

Dr. Roland is chairman of the Department of Otolaryngology--Head and Neck Surgery at the University of Texas Southwestern Medical Center in Dallas. He specializes in otology and neurotology.