What is the optimal medical therapy for necrotizing otitis externa / skull base osteomyelitis?

More specifically, the following subquestions were included:

1. What is the optimal empirical antimicrobial treatment regimen?
2. In case of failure of the initial empirical treatment, what should be changed in the empirical treatment regimen?
3. What is the optimal directed antimicrobial treatment regimen?
4. What is the optimal treatment modality? Intravenous or (partial) oral administration? Is there an added role for topical therapy?

Balance between desired and undesired effects

The most common pathogen in skull base osteomyelitis is Pseudomonas aeruginosa. Less frequent pathogens are fungi such as Aspergillus spp and Staphylococcus aureus (Takata, 2023) but a substantial portion of cases remain without identified pathogen.

Empirical therapy

Current practice is highly variable, as illustrated by the guidelines of Dutch hospitals, which can be found in a nationally summarized from here). It shows 5 different treatment strategies, ranging from oral monotherapy to combination therapy with two intravenous antibiotics. Most medical centers make use of an combination of a cephalosporin (e.g., ceftazidime) or broad-spectrum penicillin (piperacillin [with tazobactam]) in combination with an aminoglycoside (e.g., tobramycin). However, monotherapy comprising the aforementioned betalactam antibiotics without the addition of the aminoglycoside is also used.

Based on the current systematic literature search for empirical therapy, the effect of combination antimicrobial therapy compared to monotherapy on the remission outcome or any other outcomes of necrotizing otitis externa is highly uncertain as the search did not identify any studies on the subject.

Directed therapy

Directed therapy, based on susceptibility tested in the laboratory, consists preferably of an oral fluoroquinolone in the case of Pseudomonas aeruginosa, or a different directed therapy against the cultured pathogen(s). Fluoroquinolones, like ciprofloxacin or levofloxacin are the only oral medicines available against Pseudomonas aeruginosa. The other treatment regimens are only available intravenously, which makes outpatient parenteral antibiotic therapy (OPAT) a necessity for treating patients at home. A (full) intravenous treatment is not necessary in bone infections in general. Oral treatment with antimicrobials with high bioavailability, such as quinolones, has proven to be non-inferior to intravenous therapy in bone infections in general (Li, 2019) The susceptibility profile of the pathogen should drive the choice of treatment, not the fact that it can be administered intravenous or orally. A recent review showed no clear differences in effectiveness of monotherapy for these agents against severe Pseudomonas spp. infections overall (Tekes-Manuva, 2024). When a different pathogen than Pseudomonas aeruginosa is found, treatment should be adapted according to susceptibility testing and advice from the local infectious diseases expert.

Based on the current systematic literature search, the effect of combination antimicrobial therapy compared to monotherapy on the remission outcome of necrotizing otitis externa is highly uncertain. The confidence interval around the estimated effect in the single included study on directed therapy is wide and allows for the possibility of both clinically relevant benefits and clinically relevant harm. It is also highly uncertain based on the literature what the difference is in adverse effects (e.g., medication side effects) between the two strategies in the treatment of necrotizing otitis externa. It should be noted that side effects are often specific to the particular medication, so the choice of antimicrobial agents within the strategy plays an important role.

General consideration on combination therapy versus monotherapy

Given the aforementioned paucity of evidence, the committee is of the opinion that the choice between combination therapy or monotherapy for both empirical and directed therapy should be based on expert opinion based on theoretical considerations and literature from other but similar infections. These will be dicussed here.

The theoretical basis for the potentially superior effectiveness of combination therapy over monotherapy rests on two different arguments. First, it is based on the idea that therapy with two antimicrobial agents to which the micro-organism is susceptible leads to better outcomes in severe infections, such as with Pseudomonas aeruginosa, due to a synergistic effect, compared to treatment with just one active agent. Secondly, it increases the likelihood that an empirical regimen includes at least one active agent against the microorganism. However, literature from other serious infections shows no convincing evidence for the first hypothesis, for instance in the case of mortality and other important outcomes in sepsis (Sjövall, 2017; Paul, 2014) and Pseudomonas aeruginosa bacteremia or pneumonia (Onorata, 2022). This is also reflected in the recommendations for empirical and directed treatment of sepsis, where combination therapy is not advised in sepsis or septic shock unless there is a high suspicion of multidrug-resistant microorganisms (Evans, 2021). Similarly, in patients with diabetic foot infections, where bone involvement, Pseudomonas aeruginosa, and diabetes mellitus also play a significant role, combination therapy is not recommended (Lipsky, 2020).

Lessons can also be learned from the literature on other infections regarding the adverse effects of combination therapy. From this it is clear that more antibiotic use leads to more antibiotic resistance (Bell, 2014), Clostridioides difficile infections (Slimings, 2021), and other side effects (Paul, 2014; Arulkumaran, 2020).

Treatment modality

Parenteral versus oral administration

The current systematic review was not specifically focused on the optimal treatment modality (parenteral versus oral) but it did not identify any comparisons based on treatment modality. Importantly, the commonly presumed superiority of parenteral over oral therapy has no convincing pathophysiological basis and studies on bloodstream, endocarditis and bone-infections have shown similar or superior efficacy of partial or fully oral regimens versus fully intravenous regimens (Wald-Dickler, 2022; Li, 2019; Iversen 2019). The committee is therefore of the opinion that oral treatment should be the preferred administration method whenever possible, as this is associated with lower cost and fewer side effects (Wald-Dickler, 2022). Additionally, the committee has formulated the following prerequisite conditions for oral treatment based on expert opinion:

• Availability of an antimicrobial agent with good oral bioavailability* for which the pathogen is:
  • Proven to be susceptible,
  • In case of empiric treatment, the presumed pathogen is likely susceptible
• Absence of sepsis or shock
• Absence of factors prohibiting absorption of the antimicrobial agent, such as severe diarrhea or vomiting, short-bowel syndrome, concomitant administration of oral iron or magnesium suppletion etc.
• No or only limited interactions with other medication, e.g., those causing prolongation of QT time.

*oral bioavailability refers to the extent to which an orally ingested drug is absorbed into the systemic circulation

Local treatment

According to the guideline committee, the role of local treatment in necrotizing otitis

externa is limited and is not considered primary therapy. Necrotizing otitis externa is an infection which affects deeper tissues and often the bone of the skull base, requiring systemic antibiotics. However, topical treatment can be administered as an adjunctive in certain situations (Rosenfeld, 2014).

Local treatment can be administered in adjunct to systemic therapy in the following

situations:

1. To physically clean the external auditory canal and remove pus or debris.
2. To reduce local symptoms (such as otorrhea and inflammation).

Topical agents that can be used are antibacterial ear drops (with or without corticosteroids) or antiseptic solutions.

Conclusion

Empirical therapy

In the opinion of the committee, there is no compelling argument to prefer combination therapy over monotherapy, as long as there is low risk for resistance of the main expected pathogen (i.e., Pseudomonas aeruginosa) against the empirical regimen. This is based on evidence from infections other than necrotizing otitis externa and considering antimicrobial stewardship principles to minimize antibiotic selection pressure and adverse events. The probability of resistance of the expected pathogen will generally depend on the context:  previous cultures in a patient play an important role, as well as local or national degree of resistance. In many countries, this will come down to a choice for one of the commonly used anti-pseudomonal antibiotics such as ciprofloxacin, ceftazidime or piperacillin-tazobactam; or a carbapenem (e.g., meropenem) for countries with higher resistance levels.

Directed therapy

In the opinion of the committee, given the evidence from infections other than necrotizing otitis externa and considering antimicrobial stewardship principles to minimize antibiotic selection pressure and adverse events, there is no compelling argument to prefer combination therapy over monotherapy.

Treatment modality

Oral treatment is preferred wherever possible for patients with mild disease as judged by the clinician. For most cases the only suitable oral agent will be an anti-Pseudomonal quinolone like ciprofloxacin. Topical treatment cannot replace systemic antimicrobial treatment but may have an adjunctive role for specific goals (i.e., external ear canal decontamination, reduction of local symptoms and to promote local healing).

Empirical therapy in case of failure of previous regimen

In cases where initial empiric antimicrobial treatment has failed, the empiric regimen should be amended based on the most likely reasons for failure. Possible reasons for treatment failure include:

1. No or insufficient antimicrobial coverage of the causative pathogen, due to resistance (e.g., resistant Pseudomonas aeruginosa) or intrinsic non-susceptibility (e.g., a non-bacterial cause like Aspergillus spp., or ceftazidime-treatment for necrotizing otitis externa caused by Staphylococcus aureus).
2. Insufficient intestinal absorption in case of oral therapy, for instance due to diarrhea or other gastro-intestinal disease, farmacokinetic interactions or non-adherence.
3. Incorrect diagnosis, for instance in case of malignancy as cause of the symptoms.

In the absence of a clear explanation for treatment failure, the committee suggests to consider performing diagnostic investigations such as (repeat) biopsy to get more information on the diagnosis in these cases. If no additional information about the etiology is obtained (yet) the committee is of the opinion that changing or extending the antimicrobial regimen to include coverage of fungal pathogens like Aspergillus spp. should be considered. Choice of antifungal regimen will depend on the most actual information on resistance and advice from a local infectious disease expert, preferably as part of a multidisciplinary team.

Specific recommendations for empirical therapy for the Dutch situation

For the current (2025) Dutch situation, Pseudomonas aeruginosa data for hospital outpatients show resistance percentages of 11, 3 and 5%  for ciprofloxacin, ceftazidime and piperacillin-tazobactam respectively (de Greeff, 2024). Levofloxacin, which is an orally available quinolone like ciprofloxacin, is another option for oral therapy but seen as second choice behind ciprofloxacin due its lower in vitro activity and higher minimum inhibitory concentration s (MICs) against Pseudmononas (MacGowan, 1999). Ciprofloxacin also has a higher mutant prevention concentration in Pseudomonas (Hansen, 2006). Ciprofloxacin can therefore be assumed to have superior potency and a lower risk of developing of resistance during treatment of Pseudomonas infections. The opposite is the case for S. aureus; In a multicenter trial to uncomplicated skin and skin structure infections, levofloxacin achieved 100% eradication of S. aureus versus 87% for ciprofloxacin (Nichols, 1997).  Although we have to be cautious not to extrapolate these studies directly to efficacy in NOE, there does seem to be a generally different efficacy of levofloxacin and ciprofloxacin on P. aeruginosa and S. aureus and we consider ciprofloxacin the first-choice quinolone for Pseudomonas infections. Quinolones like ciprofloxaxin are the only agents where oral administration is possible. This option is not recommended for patients already treated with ciprofloxacin (or another quinolone) during the current course of disease as the probability of resistance will be increased.

For all other cases the choice is either empirical ceftazidime or piperacillin-tazobactam monotherapy. In situations where the expected resistance percentage is similar, the choice between ceftazidime and piperacillin-tazobactam can also be guided by the advantage of methicillin-susceptible Staphylococcus aureus coverage by piperacillin-tazobactam, which can be a cause of the infection in a minority of cases. Methicillin-resistant Staphylococcus aureus (MRSA) is rare in the Netherlands. On the other hand, the additional anaerobic coverage supplied by piperacillin-tazobactam may have (thusfar mainly theoretical) disadvantages due to its impact on the patient’s enteric microbiota with potential harmful clinical consequences (Kullberg, 2024).  

Based on the above, the committee is of the opinion that for the current Dutch situation empirical therapy with piperacillin-tazobactam is the first choice therapy in all severe cases and all other cases where oral ciprofloxacin treatment is not possible. Ceftazidime therapy is a good alternative choice, offering slightly lower Pseudomonas spp. resistance, but is not effective against Staphylococcus aureus which is a reported as a cause in around 6% of cases (Takata, 2023). Because of the uncertainty of the potential downside of the anaerobic coverage of piperacillin-tazobactam, the committee prefers the advantage of the additional Staphylococcus aureus in this case.

Quality of the evidence

For empirical therapy, no studies meeting our criteria could be identified. For directed therapy, the overall quality of evidence is very low. This means that we are very uncertain about the estimated effect found for the critical outcome measures. The evidence was downgraded due to:

• Risk of Bias: methodological limitations in study design.
• Imprecision: the confidence interval exceeds both limits of clinical relevance; not achieving the optimal sample size.

Moreover, the antibiotic agent, cefsulodin that was used in the only study that was included, is not often used in currently clinical practices and not available in the Netherlands.

Values and preferences of patients (and possibly their caregivers)

The primary preferred outcome of patients will be effective treatment of the infection. Given that, the possible options for therapy mainly differ with regard to the possibility for oral administration (generally only possible for ciprofloxacin) as this precludes the need for intravenous cathethers with their added discomfort (e.g., pain) and complications (e.g., infection, bleeding). Also, intravenous therapy generally necessitates hospital admission for at least a part of the therapy duration and/or outpatient parenteral therapy. The committee therefore assumes that oral therapy will be preferable for patients provided that the chance of treatment success is high.

Costs (resources)

No studies on cost or cost-effectiveness were identified. Medical costs will generally be lower for cases where all or a part of treatment can be delivered orally.

Equity ((health) equity/equitable)

The guideline panel expects no problems with health equity with regard to medical treatment of necrotizing otitis externa.

Acceptability

The guideline panel expects no problems with ethical acceptability or sustainability with regard to medical treatment of necrotizing otitis externa.

Feasibility

The intervention seems feasible. The intervention is generally already standard care in practice. For cases where no oral treatment is possible, outpatient parenteral antibiotic therapy (OPAT) would be preferable, but this is not universally available, which may lead to prolonged hospital admissions.

Further explanation about OPAT is described in module 9: additional conditions optimizing care.

Rationale of the recommendation: weighing arguments for and against the interventions

The literature on the desired and undesired effects of combination versus monotherapy in necrotizing otitis externa provides insufficient certainty, but literature and guidelines on relevant other infections provides evidence that there is no added value of combination empirical or directed therapy, but increased risks of adverse effects, including side effects and antibiotic resistance, may be expected. Therefore, the committee is of the opinion that the optimal empirical treatment regimen consists of monotherapy with ciprofloxacin or a beta-lactam antibiotic for which Pseudomonas aeruginosa is likely susceptible, which will depend on the local resistance situation and previous cultures. In some clinical contexts this may mean that treatment with multiple antimicrobials is needed to form a regimen with likely activity against Pseudomonas aeruginosa; however, this is different from combination therapy which entails giving therapy with two or more presumed active antimicrobials.  Preferred directed therapy is also monotherapy with an antimicrobial agent for which the pathogen is proven susceptible. If possible, upfront oral therapy or early intravenous-to oral switch is preferable to minimize cost and side effects without reducing efficacy, given specific conditions for oral therapy (see above).

Final judgment:

Weak recommendation for empirical monotherapy. Weak recommendation for directed monotherapy.