Cystic fibrosis: Antibiotic therapy for pulmonary exacerbations

INTRODUCTION — Pulmonary disease remains the leading cause of morbidity and mortality in people with cystic fibrosis (CF) [1]. Despite important advances in treatment, pulmonary exacerbations are a major driver of accelerated loss of lung function, decreased quality of life, and increased mortality [2-5].

Selection of antibiotics for pulmonary exacerbations is discussed below. An overview of pulmonary exacerbations is presented in a separate topic review, including clinical assessment, pathogenesis, interventions to promote secretion clearance, respiratory support, site of care, and the role of antiinflammatory and antiviral agents. (See "Cystic fibrosis: Management of pulmonary exacerbations".)

Other aspects of pulmonary disease in CF are discussed in separate topic reviews:

●(See "Cystic fibrosis: Antibiotic therapy for chronic pulmonary infection".)

●(See "Cystic fibrosis: Overview of the treatment of lung disease".)

●(See "Cystic fibrosis: Clinical manifestations of pulmonary disease".)

●(See "Cystic fibrosis: Management of advanced lung disease".)

●(See "Cystic fibrosis: Treatment with CFTR modulators".)

RATIONALE FOR ANTIBIOTIC TREATMENT — Systemic antibiotic treatment is indicated for all patients with acute pulmonary exacerbations and is recommended in virtually all consensus guidelines, in conjunction with a comprehensive regimen including promotion of airway clearance, continuation of modulator therapy (if eligible), optimization of nutrition, and management of CF-related diabetes (if present) [6-9]. Although antibiotic therapy is a nearly universal practice and reflects widespread expert opinion, it is based primarily on clinical experience and indirect evidence [10]. Pulmonary exacerbations are usually identified based on changes in symptoms and/or spirometry metrics, but the CF field has not reached a consensus definition. (See "Cystic fibrosis: Management of pulmonary exacerbations", section on 'Definition'.)

The benefit of antibiotics for treating pulmonary exacerbations in patients infected with Pseudomonas aeruginosa has been evaluated in only a small number of placebo-controlled clinical trials [11-14]. All were published between 1980 and 1990, had fewer than 30 participants per study, and were of poor quality by today's standards [15]. Some but not all reported improved outcomes in the groups receiving antipseudomonal antibiotics.

Indirect evidence of the value of antibiotic treatment for exacerbations comes from observational studies reporting that bacterial density (particularly for P. aeruginosa) is correlated with pulmonary symptoms [16] and that bacterial density and inflammatory markers decrease after antibiotic treatment of pulmonary exacerbations [17]. In a study of nearly 10,000 individuals who had an acute drop in forced expiratory volume in one second (FEV₁) of ≥10 percent from baseline, those whose episodes were diagnosed as pulmonary exacerbations and treated with antibiotics were more likely to recover to ≥90 percent of their baseline FEV₁ values [18].

Given the scarcity of high-quality evidence regarding antibiotic therapy, most recommendations are based on expert opinion, resulting in considerable variation in antibiotic-prescribing practices among CF clinicians [19]. In an effort to standardize care, the CF community has adopted guidelines that will be described below. Some of these practices are being actively challenged and are being studied in clinical trials. Decisions regarding antibiotic use need to balance their potential benefits against the burdens of cost, time for administration, and risks of adverse drug reactions.

OUR APPROACH TO ANTIBIOTIC SELECTION — We select antibiotics using an individualized approach, based on the following considerations:

●Pathogens identified in respiratory secretion cultures – We generally select antibiotics to cover each pathogenic species identified in the patient's recent respiratory secretion cultures, as recommended by virtually all guidelines [6,7,9]. Commonly used antibiotic combinations are outlined in the table (table 1). We select at least two antibiotics to cover P. aeruginosa infections, although this practice is being actively reassessed, and least one antibiotic to cover any other pathogenic gram-negative bacteria (including Achromobacter species, Burkholderia cepacia complex species, and Stenotrophomonas maltophilia) [6,7,9]. (See 'Respiratory secretion cultures' below and 'Double coverage for P. aeruginosa' below.)

●Antibiotic susceptibility testing – For most pathogens, we use the patient's most recent antibiotic susceptibility test results to select antibiotics. However, for P. aeruginosa, we select antibiotics empirically based on their established antipseudomonal properties, often prescribing the same regimen that successfully treated the previous pulmonary exacerbation, because antibiotic selection based solely on susceptibility testing of P. aeruginosa appears not to improve outcomes. It is not always possible to cover each pathogen identified in the patient's sputum or to cover some multidrug-resistant organisms. (See 'Antibiotic susceptibility testing and its limitations' below.)

●Response to previous exacerbations – We prescribe the same antibiotic regimen that was previously successful, as measured by changes in symptoms and pulmonary function tests, unless the bacterial species identified in respiratory secretions have changed since the last episode.

●Severity of the exacerbation – For mild exacerbations, we often begin with an oral regimen. For moderate or severe exacerbations or if the response to the initial oral regimen is suboptimal, we use a regimen that usually includes antibiotics that require intravenous (IV) administration. (See 'Severity of the exacerbation' below.)

●Other – Other considerations include the patient's drug allergies (which are common), toxicity risks, and patient preferences. We generally do not use inhaled antibiotics as a substitute for systemic antibiotics to treat moderate or severe exacerbations. (See 'Other patient-specific considerations' below and 'Inhaled antibiotics' below.)

This individualized approach reflects the uncertain value of using susceptibility testing to guide antibiotic selection, particularly for P. aeruginosa. (See 'Antibiotic susceptibility testing and its limitations' below.)

KEY CONSIDERATIONS

Respiratory secretion cultures — We suggest performing cultures of expectorated sputum or throat swabs for nonexpectorating patients at least every three months during routine clinic visits, consistent with guidelines from the Cystic Fibrosis Foundation [6,8,20,21] and other expert groups [7,9,22]. Because people with CF often carry the same bacteria for prolonged periods of time, recent cultures are reasonably predictive of the pathogens present in a subsequent pulmonary exacerbation.

If a routine culture has not been performed within the few weeks prior to the start of an exacerbation, we usually obtain one at that time (see "Cystic fibrosis: Antibiotic therapy for chronic pulmonary infection", section on 'Periodic surveillance cultures'). However, we modify the initially selected regimen only if the new culture results identify a new bacterial species that is not covered by the original regimen. There is insufficient information to assess the value of performing bronchoalveolar lavage to obtain fluid for culture at the time of a pulmonary exacerbation. The procedure is performed rarely in adults and occasionally in children. (See 'Response to a failing regimen' below.)

We do not broaden antibiotic coverage to cover pathogens beyond those identified in culture, because this does not improve clinical outcomes [23] and may promote acquisition of multidrug-resistant bacteria such as methicillin-resistant Staphylococcus aureus (MRSA) and multidrug-resistant P. aeruginosa [24].

Double coverage for P. aeruginosa — For P. aeruginosa, it common practice to treat pulmonary exacerbations with least two antibiotics, consistent with multiple CF guidelines [6,7,9]. The rationale for double coverage is that chronic infection with this bacterium is associated with a more rapid decline in pulmonary status and on the assumption that double coverage will have synergistic effects and might delay the generation of antibiotic resistance. However, note that there is little evidence to support the benefits of double coverage for P. aeruginosa, as summarized in a systematic review [25], subsequent retrospective studies using registry data [26], and secondary analysis of data from the STOP2 trial [27,28]. Regarding antibiotic resistance, two trials found that double therapy was associated with a trend toward a decrease in the frequency of antibiotic resistance compared with monotherapy, but the meta-analysis did not detect a statistically significant difference [25]. Fortunately, a prospective randomized study is underway to explore the utility of double antipseudomonal coverage (specifically, adding tobramycin to an intravenous [IV] antipseudomonal beta-lactam; NCT05548283).

Antibiotic susceptibility testing and its limitations

●Clinical approach – For P. aeruginosa, we select antibiotics based on their established antipseudomonal properties, the patient's response to previous regimens, and toxicity considerations; we use antibiotic susceptibility testing as a minor consideration. For other bacteria (eg, S. aureus, S. maltophilia or Achromobacter species), we are continuing the traditional practice of selecting antibiotics based on susceptibility test results, pending further study.

The traditional practice of using susceptibility test results to guide antibiotic selection for P. aeruginosa is not well supported by accumulating data and is being actively questioned, because of accumulating evidence that this practice does not improve outcomes [29-32]. In a systematic review, susceptibility results failed to predict clinical response in 11 of 13 studies [29]. Among the two remaining studies, one study found a correlation for only of two treatment regimens being evaluated [33]. In the second study, multivariable logistic regression analysis showed that the correlation could be explained by imbalances in other risk factors known to influence treatment outcomes [34]. Finally, a subsequently published retrospective study of 2390 pulmonary exacerbations in 413 patients reported that improvements in forced expiratory volume in one second (FEV₁) and body weight were no different between groups whose antibiotic selection fully covered, partially covered, or left uncovered the P. aeruginosa bacteria that were isolated [35].

This evidence has led some CF centers to decrease the frequency of routine antibiotic susceptibility testing to once a year for P. aeruginosa without apparent deleterious effect [30]. A group of experts was assembled by the Cystic Fibrosis Foundation to use a Delphi approach to generate a list of best clinical practices regarding selection of antibiotics [22]. No consensus could be reached regarding the use of antibiotic susceptibility test results to guide antibiotic selection for treatment of P. aeruginosa during acute exacerbations.

There are multiple possible explanations for the problems with antibiotic susceptibility testing as performed in clinical microbiology laboratories, including inconsistent results on serial testing [36], variable results from specimens obtained from different airway segments [37], and differences in the properties of bacteria when grown under laboratory conditions compared with how they exist within CF airways conditions. Microbiome studies have demonstrated that the spectrum of bacteria present in the CF airway is much broader than what is identified by culture techniques [38]. The role of the previously unrecognized bacteria in pulmonary exacerbations is unknown.

●Alternate strategies – To improve the utility of susceptibility testing, alternate strategies have been evaluated, but the results so far do not support their clinical use. One strategy is to culture bacteria under conditions that induce biofilm formation to mimic what occurs in the CF airway because bacteria grown in biofilms are generally less susceptible to antibiotics compared with bacteria grown under standard laboratory conditions [39]. However, clinical studies found no differences in patient outcomes or sputum bacterial density when antibiotic regimens were chosen based on susceptibility results from biofilms versus conventional cultures, as outlined in a systematic review [40]. Another strategy is to perform antibiotic synergy tests to determine if a multidrug-resistant isolate is susceptible to combinations of antibiotics, although they are resistant to each drug when tested separately [41,42]. Unfortunately, a large randomized trial showed no difference in clinical outcome when antibiotics were selected based on synergy testing compared with standard susceptibility testing [43].

●Coverage of multiple bacteria – It is not unusual for CF patients to have multiple bacterial species identified in their respiratory secretions. Selecting an antibiotic combination that covers all of the isolates is occasionally difficult without resorting to an impractically large number of antibiotics. Unfortunately, little information is available to determine the priority of the different pathogens when only a subgroup can be reasonably covered.

●Anaerobic bacteria – The role of anaerobic bacteria in pulmonary exacerbations remains uncertain [44]. Although not detected in respiratory secretions using standard culture methods, anaerobic bacteria are commonly found in CF respiratory secretions by nonculture-based methods, with their relative abundance often increasing immediately prior to pulmonary exacerbations [45]. Nonetheless, a retrospective study of 514 pulmonary exacerbations in 182 subjects found that antibiotic coverage of anaerobic bacteria did not lead to better recovery of baseline of FEV₁ or prolong the time to the next exacerbation [46]. We therefore do not select antibiotics with the specific purpose to cover anaerobic bacteria.

●Antibiotic-resistant bacteria – When in vitro testing can identify no antibiotic to which a bacterium is susceptible, our practice is to select from a list of antibiotics that would otherwise be chosen empirically for that pathogen (table 1). Retrospective studies indicate that many patients will improve clinically under these circumstances [29,31]. Furthermore, the lack of correlation between susceptibility test results and clinical outcomes for P. aeruginosa indicate that clinicians should not be hesitant to choose antipseudomonal antibiotics that the laboratory reports as resistant.

If a patient is clinically improving following initiation of antibiotics, we continue the regimen regardless of the resistance pattern reported from a sample obtained at the start of treatment. A retrospective study of 6451 pulmonary exacerbations in pediatric patients reported that antibiotic switching during treatment was more frequent when new susceptibility tests were performed but without evidence of improved outcomes [47,48].

Severity of the exacerbation — The severity of an exacerbation is typically based on the clinician's global assessment of the change in symptoms compared with the patient's baseline. An exacerbation is usually considered severe if there are substantial changes in symptoms and/or >10 percent decline in FEV₁ relative to the individual's baseline function. An exacerbation is mild if there are perceptible but small changes in symptoms or FEV₁ relative to baseline. (See "Cystic fibrosis: Management of pulmonary exacerbations", section on 'Severity grading'.)

Antibiotic selection is partly influenced by the severity of the exacerbation:

●Moderate or severe – A moderate or severe exacerbation generally warrants rapid initiation of aggressive therapy, under the assumption that doing so will increase the chances of returning to baseline status. S. aureus should be covered with a first-line drug (vancomycin, linezolid, or ceftaroline), and P. aeruginosa is typically covered with two systemic antibiotics (table 1) (see 'Our approach to antibiotic selection' above). Aggressive treatment is warranted because the risk of permanent loss of lung function if the infection persists is thought to outweigh the risks of adverse effects from drug toxicity and the burden and complications of IV access.

Note that the route by which an antibiotic is administered is determined by its bioavailability and should not be viewed as a reflection of its effectiveness. Many of the bacteria that infect CF airways require antibiotics that must be administered IV because they are not absorbed when given orally. However, certain orally administered antibiotics (ciprofloxacin and linezolid) can be included in the regimen if appropriate to the patient's pathogens; they are effective because they are highly bioavailable and achieve therapeutic levels when administered orally.

●Mild – For milder exacerbations, we agree with the common practice of using oral antibiotics to minimize the treatment burden and risk for adverse effects associated with IV treatment [48-51]. Evidence shows that this approach leads to clinical improvement [50], but the extent may not be as great compared with using more aggressive therapy such as regimens that include hospitalization and/or IV antibiotics [18,52-54]. If oral antibiotics are used, patients should be monitored closely and treatment escalated with consideration of hospitalization and/or using IV antibiotics if their inclusion better covers the bacteria identified by culture.

For some individuals with mild exacerbations, an oral regimen may not be possible if they have resistant strains of bacteria (other than P. aeruginosa, for which antibiotic susceptibility testing has limited utility) or if they have drug allergy or intolerance. (See 'Antibiotic susceptibility testing and its limitations' above.)

This practice of escalating to IV therapy for people who do not respond to oral antibiotics is supported by retrospective studies suggesting that oral antibiotics may be insufficient for some pulmonary exacerbations:

•In one study, improvement in percent predicted FEV₁ a few days after treatment initiation was better in patients receiving IV antibiotics (5.1±12.7 standard deviation) compared with those receiving oral antibiotics only (2.0±11.6) [49]. However, FEV₁ measured during the 90 days after treatment did not differ between groups.

•In another study, patients who were treated with oral antibiotics alone had slower rates of FEV₁ improvement during treatment compared with those receiving regimens that included IV antibiotics [54]. Longer-term outcomes were not reported.

Though provocative, these studies have limitations in that they are retrospective and subject to bias in treatment assignment that may not be adequately addressed by the statistical methods employed by the investigators.

Other patient-specific considerations — Although our antibiotic selection starts with the principles outlined above, other factors sometimes modify final choices based on an assessment of benefits versus risks:

●Drug allergy and toxicity considerations – Allergies to antibiotics are common in patients with CF [55] and influence the choice among the antibiotic options suggested by the culture results [55]. If acceptable alternatives are not available, desensitization protocols can be used for antibiotics that previously caused immediate hypersensitivity reactions (see "Rapid drug desensitization for immediate hypersensitivity reactions"). Any significant adverse effects that occurred during previous courses of antibiotics (eg, renal or ototoxicity from aminoglycosides) influence drug selection. (See 'Aminoglycosides' below.)

●Patient preferences – Although the treatment options we present to a patient are derived from the above considerations, patient preferences can influence the final decisions. For example, some patients have strong opinions about when they are willing to use IV antibiotics or be hospitalized in given situations. After a discussion of the benefits and risks of the different options, we incorporate the patient's preferences when forming the treatment plan.

Inhaled antibiotics — In general, when a beta-lactam and/or either an aminoglycoside or colistimethate are indicated (table 1), we deliver them parenterally and do not rely on their inhaled versions, although practice varies among clinicians [49]. (See 'Dosing considerations for people with cystic fibrosis' below.)

Evidence supporting our approach comes from a large registry study (9040 events in 3253 patients) in which the inclusion of an inhaled antibiotic did not improve outcomes, namely, increase in FEV₁ or time to next pulmonary exacerbation [56]. A previous systematic review found insufficient information from randomized trials to guide when to use inhaled antibiotics during exacerbations [57]. In addition, we suspect that inhaled antibiotics may be less effective because their distribution to the lungs of CF patients is known to be very inhomogeneous [58].

As exceptions, we might include an inhaled antibiotic in the following situations:

●For a relatively mild pulmonary exacerbation, when an inhaled antibiotic can be added to an oral medication (eg, a fluoroquinolone)

●When the inhaled antibiotic provides coverage for a particular bacterial isolate that is not otherwise covered by the chosen systemic regimen

If a patient's chronic baseline regimen includes an inhaled antibiotic (eg, tobramycin), practice varies regarding whether to suspend the inhaled antibiotic until the course of IV antibiotics is complete. (See 'Managing the chronically prescribed antibiotics' below.)

ANTIBIOTICS FOR SPECIFIC BACTERIA — Typical antibiotic regimens for each of the common pathogens in people with CF are summarized in the table (table 1) and discussed in more detail below.

●P. aeruginosa – We select antibiotics from the list of those with known antipseudomonal activity and based on the patient's response to previous regimens and toxicity considerations. We use antibiotic susceptibility testing as a minor consideration. We generally select at least two antibiotics that cover this organism, although that approach is being called into question. (See 'Antibiotic susceptibility testing and its limitations' above and 'Double coverage for P. aeruginosa' above.)

Appropriate options are (table 1):

•For mild exacerbations, we treat with an oral fluoroquinolone, either ciprofloxacin or levofloxacin, even if the clinical laboratory reports resistance, provided that the patient responded well in the recent past. Based on published pharmacokinetic studies, children with CF generally require higher doses of ciprofloxacin than other children (see 'Ciprofloxacin' below). For patients using an inhaled antipseudomonal antibiotic as part of their chronic treatment, we recommend that they continue the inhaled antibiotic or start a new course if they are in their "off" period.

•For moderate or severe exacerbations, or if the above oral/inhaled regimen fails, we treat with an antibiotic combination that includes:

-A beta-lactam – Such as piperacillin-tazobactam, cefepime, ceftazidime, ceftazidime-avibactam, ceftolozane-tazobactam, imipenem with cilastatin, or meropenem (or ticarcillin-clavulanate, where available).

Plus

-Either a fluoroquinolone (eg, ciprofloxacin or levofloxacin) or tobramycin if recent use of a fluoroquinolone has failed. We select tobramycin rather than gentamicin because tobramycin usually has greater in vitro activity against P. aeruginosa [59] and lower risk of drug-induced kidney injury [60]. We use intravenous (IV) rather than inhaled tobramycin for this purpose. (See 'Inhaled antibiotics' above.)

For patients with newly acquired P. aeruginosa, sputum cultures should be obtained following successful treatment of the exacerbation to determine if the Pseudomonas has been eradicated. If not, an "early eradication" protocol should be used. (See "Cystic fibrosis: Antibiotic therapy for chronic pulmonary infection", section on 'Prevention and eradication'.)

●Methicillin-sensitive S. aureus (MSSA)

•For mild exacerbations, we treat with trimethoprim-sulfamethoxazole, doxycycline, or amoxicillin-clavulanate when in vitro testing shows susceptibility. Based on published pharmacokinetic studies, children with CF generally require higher doses of trimethoprim-sulfamethoxazole than other children. (See 'Trimethoprim-sulfamethoxazole' below.)

•For moderate or severe exacerbations, or if the above oral regimen fails, we treat with nafcillin or cefazolin.

●Methicillin-resistant S. aureus (MRSA)

•For mild exacerbations, we treat with trimethoprim-sulfamethoxazole or doxycycline, if in vitro testing shows susceptibility to these drugs.

•For moderate or severe exacerbations, or if the above oral regimen fails, we treat with oral linezolid, IV vancomycin, or IV ceftaroline [61]. Although linezolid is administered orally, it is highly effective because of its high bioavailability and achieves therapeutic levels. Retrospective studies comparing vancomycin with either linezolid or ceftaroline have reported no differences in outcome [62,63].

Of note, S. aureus resistance to macrolides is increasing in patients treated chronically with azithromycin, causing macrolides to be less reliable for the treatment of S. aureus infections [64]. (See "Cystic fibrosis: Overview of the treatment of lung disease", section on 'Azithromycin'.)

●P. aeruginosa with MSSA or MRSA

•For mild exacerbations, we treat with an oral fluoroquinolone (either ciprofloxacin or levofloxacin) for the P. aeruginosa plus a second oral antibiotic to cover the MSSA or MRSA, as detailed above.

•For moderate or severe pulmonary exacerbations, or if the above oral regimen fails:

-When MSSA accompanies the P. aeruginosa, we treat with a combination that includes piperacillin-tazobactam, cefepime, imipenem with cilastatin, meropenem, or ticarcillin-clavulanate (the latter is not available in the United States) plus one of the following: an oral fluoroquinolone or IV tobramycin, amikacin, or colistin, as discussed above.

-When MRSA accompanies the P. aeruginosa, we treat with vancomycin or linezolid plus the same antibiotic combination as for P. aeruginosa alone (three antibiotics total). Although ceftaroline has good activity against MRSA, it is not effective for P. aeruginosa. Because we usually include an antipseudomonal beta-lactam to treat pulmonary exacerbations involving P. aeruginosa, we try to avoid ceftaroline for the MRSA out of concern for using two beta-lactams simultaneously, but we will resort to using two beta-lactams if there are no better options.

●B. cepacia complex – We attempt to treat B. cepacia complex species when present, guided by in vitro antibiotic susceptibility testing. However, this is not always possible because these bacteria (which includes Burkholderia multivorans and Burkholderia cenocepacia) are often highly resistant to multiple antibiotics. The problem of choosing an antibiotic for B. cepacia complex species is compounded by poor performance of clinical laboratory antibiotic susceptibility testing [65].

Treatment options are often limited, but some isolates show susceptibility to trimethoprim-sulfamethoxazole, doxycycline, ceftazidime, and/or meropenem. When no single antibiotic is effective, combinations of two or more antibiotics sometimes show in vitro susceptibility [42]. (See "Cystic fibrosis: Antibiotic therapy for chronic pulmonary infection", section on 'Burkholderia cepacia complex'.)

●Achromobacter – We treat Achromobacter species if present because there is evidence that some isolates can be particularly inflammatory in nature and are associated with an increased rate of forced expiratory volume in one second (FEV₁) deterioration, similar to that induced by P. aeruginosa [66-68]. Although outcome studies of antibiotics to target Achromobacter species in CF exacerbations are lacking, we recommend piperacillin-tazobactam, trimethoprim-sulfamethoxazole, meropenem, ceftazidime, or minocycline [69].

●S. maltophilia – We attempt to treat S. maltophilia when this organism is identified in patients with pulmonary exacerbations and especially for patients with a deteriorating clinical course in whom S. maltophilia is the only cultured pathogen.

Based on in vitro antibiotic susceptibilities of clinical isolates, we recommend trimethoprim-sulfamethoxazole as the primary antibiotic but recognize that resistance is increasing [69]. Alternative antibiotics are minocycline, levofloxacin, tigecycline, aztreonam/ceftazidime-avibactam, or aztreonam/amoxicillin-clavulanate.

Although we generally include S. maltophilia coverage in our antibiotic selection, evidence is conflicting regarding the importance of treating this organism. S. maltophilia is often found in the presence of other CF pathogens and is more frequent in patients with advanced lung disease [70-72]. Moreover, in our clinical experience, we have seen occasional patients with deteriorating clinical course in whom S. maltophilia is the only cultured pathogen. On the other hand, two studies suggest that acquisition of S. maltophilia may not affect subsequent disease progression, because the rate of FEV₁ decrease following new acquisition of S. maltophilia was no different from matched control patients [70,72].

●Aspergillus and Candida – The prevalence of Aspergillus species in respiratory secretions has been reported to be between 12 and 35 percent [73,74]. Although there remains conflicting information whether Aspergillus contributes to the progression of CF lung disease (see "Cystic fibrosis: Antibiotic therapy for chronic pulmonary infection", section on 'Aspergillus species'), there is no definitive evidence that Aspergillus causes pulmonary exacerbations. As such, we do not initiate antifungal therapy for pulmonary exacerbations when Aspergillus is identified. Candida species, particularly Candida albicans, are also frequently identified in CF respiratory secretions. Similar to Aspergillus, the clinical significance of their presence is doubtful [75] and we do not treat them.

DOSING CONSIDERATIONS FOR PEOPLE WITH CYSTIC FIBROSIS — Dosing recommendations for antibiotics are summarized in the table (table 1) and detailed below. The pharmacokinetics of many antibiotics are different in people with CF compared with the general population [76,77]. People with CF typically require higher and/or more frequent dosing for hydrophilic drugs (such as penicillins, cephalosporins, and aminoglycosides) due to increased volume of distribution and total body clearance.

Beta-lactams — For most beta-lactam antibiotics, we suggest increased doses for people with CF, as shown in the table (table 1).

This strategy recognizes that people with CF often require higher doses of these drugs to maintain tissue concentrations above the minimum inhibitory concentration (MIC) through much of the dosing interval. For many of these antibiotics, our suggested doses are consistent with a comprehensive review of antipseudomonal antibiotics in CF [78]. For a few of the antibiotics (piperacillin-tazobactam, ceftazidime, and ticarcillin-clavulanate), the review recommends much higher doses for people with CF than are used for the general population. These are based primarily on pharmacokinetic/pharmacodynamic considerations, but clinical trial data for these very high doses are sparse [79]. Therefore, we suggest intermediate doses for these drugs, as shown in the table. Higher doses may be considered on a case-by-case basis.

Prolonged or continuous infusion of beta-lactam antibiotics is used by an increasing number of CF programs, based primarily on theoretical considerations and limited clinical data [80]. The rationale is that prolonged/continuous infusions can better maintain blood levels above the MIC of susceptible bacteria than repeated short-duration infusions (see "Prolonged infusions of beta-lactam antibiotics"). Pharmacokinetic modeling and blood level measurements in people with CF confirm that continuous infusions achieve more prolonged time above MIC than intermittent dosing [77,79,81]. Scant data are available to determine if prolonged/continuous infusion improves clinically relevant outcomes in CF exacerbations [82]. The largest reported clinical trial compared intermittent with continuous infusion of ceftazidime [33]. This randomized crossover study enrolled 70 people with CF, but only 49 completed both arms of the trial. There was no difference in forced expiratory volume in one second (FEV₁) improvement between infusion strategies using an intention-to-treat analysis, but the per-protocol analysis showed that continuous infusion increased percent predicted FEV₁ by 4 points more than the intermittent arm. A retrospective analysis of 22 patients treated with beta-lactam antibiotics for 43 pulmonary exacerbations reported that FEV₁ recovery was not better in the subgroup of patients whose antibiotic dosing met predefined pharmacokinetic targets [83]. However, of note, no studies have reported worse outcomes using prolonged/continuous infusion.

Aminoglycosides — Starting doses of aminoglycosides for people with CF are higher than recommended for the general population (table 1) [84]. The dose and frequency from a previous course of treatment may be used initially if serum concentrations were in the target range and creatinine clearance is not substantially changed, but drug levels should still be monitored. (See "Dosing and administration of parenteral aminoglycosides".)

The bactericidal effect of aminoglycosides is concentration dependent, and they have a postantibiotic effect (ie, bacterial growth remains suppressed for several hours after the antibiotic level drops below the MIC) (see "Aminoglycosides"). People with CF tend to require higher doses of aminoglycosides than the general population because of both increased volume of distribution and accelerated renal clearance rate [85].

●Once-daily dosing (if renal function is normal) – For people with CF and normal renal function, we suggest dosing aminoglycosides once daily (known as "consolidated dosing" or "extended-interval dosing"). This approach is consistent with the guidelines endorsed by the Cystic Fibrosis Foundation [6] and is supported by a few small clinical trials and a meta-analysis [86]. This practice has been adopted by the majority of CF centers, including pediatric programs [87]. The starting dose for tobramycin is 10 mg/kg/24 hours for children and adults without renal insufficiency (table 1), although a higher dose of 13 mg/kg/24 hours has been suggested for children younger than five years of age based on pharmacokinetic modeling [88]. In patients with obesity, the initial mg/kg dose should be altered by using adjusted body weight, which is calculated as lean body weight plus 40 percent of body weight in excess of lean body weight [89,90]. (See "Dosing and administration of parenteral aminoglycosides".)

In a meta-analysis of three trials (289 participants), patients treated with once-daily aminoglycoside dosing had similar improvements in pulmonary function compared with multiple-daily dosing (difference in mean change in FEV₁ 0.3 percent, 95% CI -2.8 to 3.5) [86]. Once-daily therapy was associated with a lower risk of nephrotoxicity among pediatric patients (mean percent change in creatinine 8.2 percent lower, 95% CI 1.1-15.3), while the difference was not statistically significant in adult patients (mean difference 3.2 percent higher, 95% CI -1.8 to 8.3). The incidence of ototoxicity was low in both groups (0.4 versus 0.8 percent; relative risk 0.56, 95% CI 0.04-8.0). Of note, most of the patients enrolled in these trials were ≥5 years of age.

Extended-interval dosing in populations without CF is discussed separately. (See "Dosing and administration of parenteral aminoglycosides", section on 'Comparing extended-interval and traditional intermittent dosing'.)

•Initial dose adjustment – We do not use published tables or nomograms for selecting and adjusting aminoglycoside doses and intervals in patients with CF, because the pharmacokinetics differ from those in non-CF patients and may result in suboptimal aminoglycoside concentrations [91,92].

Instead, we measure serum levels twice following the first dose (eg, at 2 and 10 hours after the dose) and use pharmacokinetic analysis to calculate the peak serum level and to extrapolate forward to determine serum levels approximately 18 hours after the dose. Consultation with a clinical pharmacist skilled in pharmacokinetic-based drug management may be helpful and is suggested. The targets are:

-Calculated peak serum level between 20 and 30 mcg/mL for tobramycin and between 80 and 120 mg/L for amikacin [78,93]. Note that this is the estimated peak level calculated from a pharmacokinetic analysis and not the measured level from blood samples drawn early after antibiotic infusion. Samples taken less than two hours postinfusion are still within the drug distribution phase, so calculations based on them would yield incorrect estimates of peak antibiotic concentrations and clearance rates, leading to inappropriate dosing regimens [94].

-Calculated serum level ≤0.5 mcg/mL at 18 hours, so that there is at least a six-hour period prior to the next dose when the patient will have low serum levels, to minimize toxicity. Slightly higher 18-hour levels (eg, ≤1.0 mcg/mL) are also acceptable, provided that the patient's renal function and clinical status are stable and the 18-hour level will be rechecked within three to four days.

If a dose that achieves the target peak level leads to too high a serum level at 18 hours, the dosing strategy is changed to the "conventional" approach to reduce the risk of toxicity.

•Subsequent monitoring and dose adjustment – After the initial dose has been established, we measure serum aminoglycoside levels once or twice per week, with each measurement timed for several hours prior to the next dose (ie, at 18 hours following the previous dose). The appropriate frequency of monitoring depends on baseline renal function, the concomitant use of potentially renal toxic drugs, and whether the patient has a history of prior aminoglycoside toxicity. The goal is to ensure that the aminoglycoside level remains relatively low for several hours prior to the next dose (ideally ≤0.5 for tobramycin at 18 hours). An increasing 18-hour level suggests the possibility of renal injury and should prompt dose adjustment. We believe that the 18-hour time point is preferable to a true trough at 24 hours because in CF patients with normal renal function, the drug concentration is frequently below the level of detection at 24 hours, which would prevent early detection of renal impairment, as manifested by increasing drug levels. Interpreting these low serum levels can be confounded if the patient is also receiving inhaled tobramycin. As an example, serum levels one hour after inhaling 300 mg tobramycin are 1.05±0.67 mcg/mL (mean±standard deviation) [95].

In addition, we suggest measuring two levels (eg, at 2 and 10 hours after the dose) following any substantial change in dose to allow pharmacokinetic calculations and assure that targeted levels are achieved. Two time point measurements are also recommended if large changes in volume of distribution are likely to have occurred during the course of treatment (eg, sepsis with capillary leak or right-sided heart failure), although these scenarios are uncommon in CF patients having a typical pulmonary exacerbation [91].

●Conventional dosing (for selected patients) – For patients with renal insufficiency or evidence of delayed aminoglycoside clearance, we do not use once-daily dosing for aminoglycosides. Instead, we use a conventional approach based on peak and trough levels to target drug levels, as follows:

•Peak serum concentration 8 to 12 mcg/mL for tobramycin or 20 to 30 mcg/mL for amikacin (measured 30 to 45 minutes after the dose is given)

•Trough serum concentration ≤2 mcg/mL for tobramycin and <10 mcg/mL for amikacin (measured just before the next planned dose)

These are the targets used for conventional dosing of aminoglycosides, but patients with renal insufficiency will require lower doses and/or longer dosing intervals than are used for patients with normal renal function.

In this situation, the dose and frequency from a previous course of treatment may be used initially if the creatinine clearance is not substantially changed and serum concentrations were within the target range [96]. If there is no reliable historical information, we suggest consulting with an expert pharmacist to guide dosing with a pharmacokinetic analysis. If pharmacist consultation is not available, it is reasonable to use an empiric loading dose of 3.3 mg/kg (if the patient is overweight, use ideal body weight or dosing weight) [6,93] and select an initial maintenance dose and dosing interval based on the patient's creatinine clearance (table 2).

Once the initial dose and interval are established, we measure peak and trough levels once or twice per week and continue to adjust the dose and interval to ensure that target peak and trough concentrations are achieved and maintained. Blood urea nitrogen (BUN) and creatinine are measured at the same time to monitor for renal toxicity. These steps are detailed in a separate topic review. (See "Dosing and administration of parenteral aminoglycosides".)

●Aminoglycoside toxicity – Individuals exposed to aminoglycosides require close monitoring to mitigate risk for renal injury and ototoxicity.

•Monitoring during treatment – Aminoglycoside levels should be monitored closely during treatment [97-100]. The timing and frequency of monitoring depends on whether a once-daily or conventional dosing strategy is used, as discussed above. In addition, BUN and creatinine levels should be measured whenever aminoglycoside serum levels are assessed.

•Ongoing monitoring – We also suggest the following ongoing steps for people who are exposed to aminoglycosides:

-Serum magnesium - Measure serum magnesium levels periodically in patients who have received multiple aminoglycoside courses within the past year. These individuals are therefore at increased risk for isolated tubular damage, manifested by magnesium wasting.

-Hearing tests – Perform hearing tests, as recommended in a guideline from the Cystic Fibrosis Foundation [101]: as a baseline for all children and adults with CF, annually for those exposed to ototoxic medications (eg, aminoglycosides), and following each course of intravenous (IV) ototoxic medications for those with established hearing loss.

When clinically significant renal damage or ototoxicity is noted, efforts should be made to minimize subsequent use of IV aminoglycosides [101].

•Prevalence – Acute nephrotoxicity is common during aminoglycoside treatment for CF exacerbations [98,102]. For example, a retrospective study of 66 patients receiving >5 courses of aminoglycosides reported that acute kidney injury, defined by the Kidney Disease: Improving Global Outcomes serum creatinine criteria, occurred during at least one course of treatment in 50 (76 percent), most of which were of mild severity [102]. The consequences of repeated courses of aminoglycosides on renal function is less certain. Some studies have shown a correlation between number of courses of IV aminoglycosides and chronic renal impairment [97,103], while others have not [102,104]. Some of the discrepancy may be that those studies that reported chronic renal impairment collected data from when patients were more likely to be treated with gentamicin rather than tobramycin and when multiple daily doses were used rather than once daily. Patients receiving multiple courses of aminoglycosides are also at risk for developing magnesium wasting without azotemia [105].

Decreased auditory perception following a single course of IV tobramycin has been reported in patients with CF treated for a pulmonary exacerbation in two prospective observational studies [106,107]. Although the extent of hearing loss from a single course of IV aminoglycoside is generally small, the accumulative effects of multiple exposures can be quite large. A retrospective study of 165 patients at a single CF center reported that 38 percent of those receiving 1 to 10 courses of IV aminoglycoside had measurable hearing loss and 80 percent of those receiving >10 courses [108]. Another retrospective study of 106 pediatric patients using audiometry to measure hearing acuity confirmed hearing loss in those receiving multiple courses of IV tobramycin, but the magnitude of hearing loss was considered clinically imperceptible [109].

Colistin — IV colistin (colistimethate sodium [CMS]) is a useful option for P. aeruginosa strains that fail to respond to aminoglycosides and fluoroquinolones. We usually use it in combination with a beta-lactam antibiotic. Renal toxicity occurs at a frequency similar to or less than that of tobramycin (NCT02918409, unpublished results) [110,111]. We monitor renal function at least twice weekly during treatment. Colistin should not be used in combination with IV aminoglycosides, due to additive renal toxicities. Dose-related neurotoxicity manifesting as perioral and tongue tingling, paresthesia of the extremities, slurred speech, and vertigo/dizziness is relatively common and usually diminishes following dose reduction [112]. Colistin overdose can cause neuromuscular blockade.

There is potential for confusion when choosing drug doses due to variability in how the antibiotic is labeled [113,114]. Careful attention to the details of the licensed prescribing information is advised. In the United States, each vial of CMS is labeled as containing 150 mg of colistin-base activity (CBA), which is the equivalent of 4,500,000 international units (or 4.5 million units) of CMS. To convert, 1 mg CBA (United States product) = approximately 30,000 international units CMS (European Union product). We administer 2.5 to 5 mg/kg per day CBA (approximately 75,000 to 150,000 international units/kg per day CMS) divided into three doses to a maximum of 300 mg per day CBA (approximately 9,000,000 international units per day CMS) (table 1). Patients with obesity should be dosed by ideal body weight. The dosing frequency of every eight hours is supported by studies in CF patients [115] but differs from the recommendations of a consensus guidelines panel that addressed colistin use in patients with severe infections but that excluded reviewing studies in people with CF [116].

Neurotoxicity, which appears to be dose related and transient, is a common adverse effect of systemic colistin therapy in people with CF, even among those treated with the moderate doses described above [111]. As a result, we do not use the higher doses (eg, up to 7 mg/kg per day CBA and/or maximum doses of 360 mg CBA per day) that are sometimes used for non-CF populations and that have been advocated by some authors for CF [114,117]. (See "Polymyxins: An overview".)

Consensus guidelines for non-CF patients recommend polymyxin B in preference over colistin to treat invasive infections, based on superior pharmacokinetic characteristics and decreased nephrotoxicity [116]. For people with CF, clinicians have been hesitant to switch to polymyxin B because of the limited pharmacodynamic/pharmacokinetic studies of polymyxin B in this population [118,119]. A study comparing polymyxin B with colistin in people with CF did not find less renal toxicity with polymyxin B [117]. Another study reported frequent neurotoxicity at polymyxin B doses recommended for non-CF patients [119].

Vancomycin

●Initial dosing – We use the same vancomycin dose and target blood levels that are used for treating a serious pulmonary infection in people without CF because the pharmacokinetics of vancomycin are similar in people with CF compared with the general population [120].

For individuals with normal renal function, we start with a weight-based dose of vancomycin of 45 to 60 mg/kg per day in three divided doses (up to 4 g per day) for adults [121] and 60 mg/kg per day (up to 3.6 g per day) in three or four divided doses for children (table 1). Higher doses may be needed in younger children [122]. Some CF centers use protocols that include loading doses for adult patients. (See "Vancomycin: Parenteral dosing, monitoring, and adverse effects in adults" and "Staphylococcus aureus in children: Overview of treatment of invasive infections".)

Of note, a beta-lactam other than piperacillin-tazobactam should be selected when being used in combination with vancomycin and tobramycin to reduce the risk of renal toxicity [123].

●Dose-adjustment strategies – There are two methods of therapeutic monitoring for vancomycin: trough-guided dosing and area under the curve (AUC)-guided dosing, which requires the assistance of a clinical pharmacist (table 3). Our preferred approach is the AUC method, but this may vary from facility to facility. AUC-guided dosing was endorsed in a consensus guideline published jointly by four professional infectious disease and pharmacist societies [122,124,125].

•Trough-directed dosing – Measuring trough levels has been the previous standard method for adjusting vancomycin dose and continues to be so if a clinical pharmacist is not available to perform the calculations for AUC monitoring or when renal function is rapidly changing. It is also the method preferred by some experts for vancomycin dosing in children [124,125]. We usually obtain vancomycin trough concentrations immediately before the third or fourth dose after initiating vancomycin or following a dose change. We aim to achieve trough concentrations of 15 to 20 mcg/mL for adults and 7 to 10 mcg/mL for children. In observational studies in children, trough concentrations ≥15 mcg/mL have been associated with increased risk of acute kidney injury and have not been associated with improved outcomes [126-128]. If trough levels are outside of the target range, the amount of vancomycin administered with each dose can be adjusted, which will proportionally alter trough levels and AUC. Alternatively, the interval between doses can be adjusted, ideally with input from a clinical pharmacist [122].

•AUC-guided dosing – In the AUC-guided approach, the vancomycin dose is adjusted based on the ratio of the AUC over 24 hours to the MIC (AUC/MIC) [122]. Clinical pharmacists can use software programs to estimate the AUC/MIC from two serum levels taken at the post-distributional peak (one to two hours after the end of infusion) and within 30 minutes prior to the next infusion. Monitoring should begin within 24 to 48 hours of initiation of treatment. The target AUC/MIC is 400 to 600, assuming a vancomycin MIC of 1 mg/L. Additional details of AUC-guided dosing for children are provided in the consensus guideline [122].

Once the target dosing has been achieved and if renal function is stable, we monitor trough levels every 7 to 10 days. More frequent monitoring is indicated for patients with changing renal function.

Evidence supporting AUC-guided dosing includes a retrospective study of people with CF (113 adults and 42 children) that compared the outcomes of those whose dose was guided by AUC with those guided by trough levels [129]. The proportion of adults whose FEV₁ returned to baseline was higher in those who had doses guided by AUC (86 percent) than those guided by trough (57 percent). The frequency of acute kidney injury did not differ between groups, but the small number of patients who had higher grades of kidney injury were all in the trough-guided group. (See "Vancomycin: Parenteral dosing, monitoring, and adverse effects in adults" and "Staphylococcus aureus in children: Overview of treatment of invasive infections".)

Ciprofloxacin — Among the fluoroquinolones, ciprofloxacin has the best in vitro activity against P. aeruginosa. The pharmacokinetics of ciprofloxacin in people with CF are more variable than in the general population and may be altered by disease severity, concurrent drug therapy, and patient age [130-133]. Children with CF generally require higher doses of ciprofloxacin than other children. As an example, in a group of children with CF treated for severe pulmonary infection, clearance of ciprofloxacin was two times higher than in children without CF [130]. Oral is preferred over IV administration because it is well absorbed, has lower cost, and is more convenient [133].

●For children with CF, we use oral ciprofloxacin at a dose of 40 mg/kg per day (up to 2 g daily) divided every 12 hours instead of standard doses of ciprofloxacin [132,134]. Ciprofloxacin has high bioavailability and achieves therapeutic levels when administered orally. If IV administration is needed, the dose is 30 mg/kg per day (up to 1.2 g daily) in three divided doses.

●For adults with CF, we use ciprofloxacin at 750 mg by mouth twice daily as is recommended for non-CF patients with severe respiratory tract infections because the pharmacokinetics of ciprofloxacin in adults with CF appear to be similar to that of adults without CF [135,136]. Higher doses (eg, 1 g by mouth twice daily) may also be appropriate based on theoretical considerations of pharmacokinetics and the level of susceptibility of the bacteria [131].

Trimethoprim-sulfamethoxazole — For people with CF, we use doses of 160 mg trimethoprim with 800 mg sulfamethoxazole, given three times daily (rather than twice daily, as is typical for the general population).

Although pharmacokinetic information regarding trimethoprim-sulfamethoxazole in people with CF is very limited, we suggest this increased frequency of dosing, based on two studies in adolescences and adults with CF that found increased clearances of trimethoprim and sulfamethoxazole in this population [137,138].

Managing the chronically prescribed antibiotics

●Chronic inhaled antibiotics – Although practice varies, we generally suspend inhaled antibiotics during the course of IV antibiotics for a pulmonary exacerbation. Moreover, we do not use an inhaled antibiotic as a substitute for one of the antipseudomonal antibiotics that are recommended in the table (table 1). If both inhaled and IV tobramycin are used concurrently, the inhaled drug can cause a modest increase in serum levels, possibly interfering with pharmacokinetic analyses and causing errors in dosing [6]. (See "Cystic fibrosis: Antibiotic therapy for chronic pulmonary infection", section on 'Inhaled tobramycin'.)

There is no evidence that adding or continuing an inhaled antibiotic is beneficial for treating a pulmonary exacerbation. A guidelines committee of the Cystic Fibrosis Foundation could not reach a conclusion regarding the risks and benefits of administering the same antibiotic by both IV and inhaled routes [6]. A retrospective study completed after publication of the guidelines found no added benefit from adding inhaled antibiotics to an IV regimen [56]. (See 'Inhaled antibiotics' above.)

●Chronic azithromycin – We continue administering oral azithromycin during the acute exacerbation if it is a component of the chronic pulmonary regimen, with some important exceptions:

•We temporarily stop azithromycin if continuation might cause adverse interactions with the antibiotics added to treat the exacerbation (eg, risk for QTc prolongation when used with fluoroquinolones) but recognizing that the half-life of azithromycin is long.

•If the intended antibiotic regimen will include tobramycin (inhaled or IV), it is unclear whether azithromycin should be suspended. Concern has arisen that patients with acute pulmonary exacerbations may respond less favorably to tobramycin if they are receiving chronic azithromycin therapy [139,140]. A proposed mechanism is the induction of bacterial efflux pumps by azithromycin that reduces bacterial tobramycin levels [141]. There is no consensus within the CF community on how to respond to this provocative, but as yet inconclusive, information [142]. Options include discontinuing chronic azithromycin for patients who are likely to be prescribed tobramycin in the near future, continuing azithromycin but selecting antibiotics other than tobramycin to treat pulmonary exacerbations, or continuing the current practice of prescribing both azithromycin and tobramycin while waiting for more definitive data. This issue is discussed separately. (See "Cystic fibrosis: Overview of the treatment of lung disease", section on 'Azithromycin'.)

•Patients with nontuberculous mycobacteria should not be given azithromycin as their only antimycobacterial antibiotic to limit development of antibiotic resistance to it. (See "Cystic fibrosis: Overview of the treatment of lung disease", section on 'Azithromycin'.)

ONGOING MANAGEMENT

Duration of treatment

●Intravenous (IV) antibiotics – For most patients treated with IV antibiotics for a pulmonary exacerbation, we recommend that the duration of therapy be based on each patient's initial response to treatment and not by a predetermined treatment plan, as follows [27]:

•If the response to treatment is rapid (eg, ≥8 point improvement in forced expiratory volume in one second (FEV₁) and improved symptoms within 7 to 10 days of starting antibiotics), stop antibiotics after 10 days

•If the response is slower, complete a 14-day course of antibiotics

A longer course of antibiotics may be warranted for patients requiring intensive care unit (ICU) care and those who experience an acute pulmonary exacerbation despite a recent course of IV antibiotics. In such patients, antibiotics are continued until symptom and FEV₁ improvement have plateaued; typical treatment duration is 14 to 21 days.

Our suggested approach for IV treatment is supported by results from the STOP2 trial, a randomized study of 982 adult patients who were treated with IV antibiotics for a pulmonary exacerbation based on their CF clinicians' clinical assessment [27]. Patients were assessed on day 7 to 10 of treatment and were categorized as having an early "robust" response (ie, increase in FEV₁ by >8 percent predicted and decease in the Chronic Respiratory Infection Symptom Score by >11 points) or a slower response (ie, not meeting the definition of "robust" response). Early robust responders (n = 277) were randomly assigned to stop treatment after either 10 or 14 days; slower responders (n = 705) were randomized to stop treatment after either 14 or 21 days. Two weeks after completing treatment, patients who received shorter antibiotic courses had a similar rate of treatment failure compared with those who received longer courses (early robust responders 1.8 versus 3.7 percent; difference -1.9 percent, 95% CI -7.5 to 3.3; slower responders 4.5 percent versus 5.1 percent; difference -0.6 percent, 95% CI -5.0 to 2.7). In addition, there were no important between-group differences for FEV₁ change from baseline: Among early robust responders, the mean FEV₁ change was 12.8 percent in the 10-day group versus 13.4 percent in the 14-day group (difference -0.7 percent, 95% CI -3.3 to 2.0). Among slower responders, the mean FEV₁ change was 3.4 percent in the 14-day group versus 3.3 percent in the 21-day group (difference 0.1 percent, 95% CI -1.1 to 1.3). Finally, the improvements in symptom scores and incidence of drug-induced toxicity did not differ between treatment groups.

Although this clinical trial did not include pediatric patients, it is reasonable to expect that a shorter course of IV antibiotics would result in similarly comparable outcomes in this population. Thus, we suggest using the same approach in both children and adults. This clinical trial also did not include patients who require ICU care and those who experience a CF exacerbation despite a recent course of IV antibiotics. A relatively long course of antibiotics (eg, 14 to 21 days) may be appropriate for such patients.

These recommendations may represent a change in clinician prescribing habits and patient expectations at many CF centers. Our experience is that a small but significant number of patients with moderate to severe exacerbations who do not return to their baseline symptom level by 14 days strongly request prolongation of antibiotic treatment. However, the clinical evidence cited above suggests that extending treatment is unlikely to achieve additional benefit.

●Oral antibiotics – Oral antibiotics are usually prescribed for 14 to 21 days [48,51,143]. It is uncertain whether the results of the STOP2 trial [27], which studied patients receiving IV antibiotics, most of whom had some time in hospital, can be extrapolated to patients receiving all oral regimens that are usually delivered at home. Response to outpatient treatment may be slower than inpatient treatment, where adequate rest, airway clearance therapy, nutrition, and on-time delivery of medications are likely better. (See "Cystic fibrosis: Management of pulmonary exacerbations", section on 'Home management of exacerbations' and "Cystic fibrosis: Management of pulmonary exacerbations", section on 'Site of care'.)

Prognosis after recovery from a pulmonary exacerbation is discussed separately. (See "Cystic fibrosis: Management of pulmonary exacerbations", section on 'Prognosis'.)

Response to a failing regimen — Lack of improvement during antibiotic treatment should prompt a reexamination for contributing factors such as an asthmatic component to the exacerbation or the presence of other pathogenic organisms, eg, a virus, fungus, or mycobacterium. If a culture obtained at beginning of treatment reveals a new bacterial species not covered by currently administered antibiotics such as methicillin-resistant S. aureus (MRSA), we change the antibiotic regimen to treat it. Although this is a common practice if a patient is not improving as expected, two observational studies did not detect benefits of switching antibiotics [47,144]. Data are not available to assess the utility of bronchoalveolar lavage to guide a change in antibiotics when a patient fails to meet treatment goals. The procedure is performed more frequently in pediatric than adult patients and is not without complications. However, with fewer patients producing sputum following introduction of highly effective modulator therapy, the role of bronchoalveolar lavage after a failed treatment course is worthy of consideration.

Of note, prolonging treatment beyond two weeks did not lead to better symptom outcome or FEV₁ recovery in a large, randomized clinical trial [27]. (See 'Duration of treatment' above.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Cystic fibrosis".)

SUMMARY AND RECOMMENDATIONS

●Overview – Antibiotic therapy plays a central role in managing acute pulmonary exacerbations in patients with cystic fibrosis (CF). This practice is based on extensive clinical experience and observations that bacterial density (particularly Pseudomonas aeruginosa) correlates with pulmonary symptoms and that bacterial density and inflammatory markers decrease with antibiotic treatment. (See 'Rationale for antibiotic treatment' above.)

Other aspects of managing pulmonary exacerbations are discussed separately, including clinical assessment, interventions to promote secretion clearance, respiratory support, site of care (home, hospital, or intensive care), and the role of antiinflammatory and antiviral agents. (See "Cystic fibrosis: Management of pulmonary exacerbations".)

●Antibiotic selection – Usual practice is to select at least one antibiotic to cover each bacterial isolate that is cultured from respiratory secretions. Acceptable antibiotic choices for common pathogens are summarized in the table (table 1). The choice is also influenced by the patient's response to prior regimens, severity of the exacerbation, and antibiotic susceptibility testing. (See 'Our approach to antibiotic selection' above.)

For most patients, we suggest double rather than single antibiotic coverage for P. aeruginosa infections (Grade 2C). However, this issue is actively being reexamined and this recommendation may change as more data become available. (See 'Respiratory secretion cultures' above and 'Double coverage for P. aeruginosa' above.)

●Route of administration

•Moderate to severe exacerbations – For patients with moderate to severe exacerbations (eg, substantial changes in symptoms and/or >10 percent decline in forced expiratory volume in one second [FEV₁] relative to the individual's baseline function), we suggest intravenous (IV) rather than oral or inhaled antibiotics unless a highly effective oral antibiotic such as linezolid is available for the targeted pathogen(s) (Grade 2C). Another exception is when an inhaled antibiotic provides coverage for a particular bacterial isolate not covered by the IV agent(s). If the chronic regimen includes an inhaled antibiotic (eg, tobramycin), we generally suspend the inhaled antibiotic until the course of IV antibiotics is complete, although practice varies. (See 'Severity of the exacerbation' above and 'Inhaled antibiotics' above and 'Managing the chronically prescribed antibiotics' above.)

•Mild exacerbations – For patients with mild pulmonary exacerbations, we suggest an oral antibiotic (eg, a fluoroquinolone) plus an inhaled antibiotic (Grade 2C). The exceptions are when the patient has a resistant strain of bacteria not susceptible to an oral agent and when the patient has an allergy or intolerance to the oral agent. (See 'Severity of the exacerbation' above and 'Ciprofloxacin' above and 'Inhaled antibiotics' above.)

●Antibiotic dosing for people with CF

•The pharmacokinetics of many antibiotics differs in patients with CF compared with normal individuals. People with CF generally require larger and/or more frequent dosing for penicillins, cephalosporins, trimethoprim-sulfamethoxazole, and fluoroquinolones. (See 'Dosing considerations for people with cystic fibrosis' above.)

•For aminoglycosides, starting doses should be larger than those recommended for individuals without CF, but dosing must be adjusted based on pharmacokinetic analysis of serum levels because of considerable interindividual variation in clearance rates. For CF patients with normal renal function, we suggest once-daily aminoglycoside dosing ("consolidated dosing") rather than conventional dosing and monitoring, with adjustments of dose and timing based on monitoring of drug levels (Grade 2B). Once-daily dosing has comparable efficacy with conventional dosing and monitoring but has advantages of possibly reducing the risk of nephrotoxicity and simplifying administration and monitoring. (See 'Aminoglycosides' above.)

●Antibiotic duration – We suggest that the duration of IV antibiotic therapy be based on the initial response to treatment, as follows (see 'Duration of treatment' above):

•For patients with a rapid response to treatment (eg, ≥8 point improvement in FEV₁ and improved symptoms within 7 to 10 days of starting IV antibiotics), we suggest a 10-day course of antibiotics (Grade 2B)

•For patients with a slower response, we suggest a 14-day course of antibiotics (Grade 2B)

A longer course of IV antibiotics may be warranted for patients requiring intensive care unit (ICU) care and those who experience a CF exacerbation despite a recent course of IV antibiotics. In such patients, antibiotics are continued until symptom and FEV₁ improvement have plateaued (typical duration is 14 to 21 days).

For most patients treated with oral antibiotics, we suggest a 14- to 21-day treatment course (Grade 2C), with escalation of therapy if clinical goals are not met.