The proportion of patients requiring at least one course of IV antibiotics per year has not decreased significantly over the past decade ¹ . Although this might seem concerning in light of the overall improvements in lung function observed over the same time period, this appears to be driven by a lower threshold among clinicians to diagnose and treat PEx ⁸ . Some important studies have highlighted the wide variability in the recognition and treatment of PEx between CF clinics and individual CF physicians in both the US and the UK ^{9, 10} . A multi-center study in the UK demonstrated higher IV antibiotic use among centers with higher baseline FEV ₁ % predicted ¹⁰ , whereas an older multi-center US study found variability among centers with respect to care (those with a higher median FEV ₁ % predicted had more frequent monitoring and antibiotics) ⁹ . These findings have led to quality improvement initiatives to reduce variability ^{11, 12} .

In order to prevent PEx and their associated sequelae, there has been a great deal of interest in identifying the risk factors for future PEx. VanDevanter et al. investigated factors associated with increased risk of PEx requiring IV antibiotics ¹³ . The study found that, out of numerous clinical variables, including sputum microbiology and treatment characteristics, the strongest risk factor for a PEx requiring IV antibiotic therapy was the occurrence of a PEx requiring IV antibiotics in the preceding year ¹³ . Not surprisingly, individuals with three or more exacerbations had the highest risk of future PEx compared with those with one or two exacerbations ¹³ .

With regard to PEx treatment outcomes, exacerbations treated with oral antibiotics are often considered to be milder events, whereas those treated with IV antibiotics are considered to represent more severe events. However, labeling a PEx on the basis of route of antibiotic treatment (oral versus IV) is limited and potentially biased, since factors other than PEx severity (based on symptoms, inflammatory markers, or lung function decline or a combination of these) may influence the decision of antibiotic route. For instance, IV antibiotics may be chosen on the basis of medication allergies/intolerances, bacterial resistance to oral antibiotics, or other non-disease severity-related factors (such as psychosocial issues or insurance coverage).

In support of the concept that exacerbations treated with oral antibiotics may not represent mild events, based on a recent retrospective study using the Toronto CF database from 2000 to 2014, nearly 20% of exacerbations treated with oral antibiotics did not recover to within 90% of baseline FEV ₁ % predicted within 3 months of treatment ¹⁴ . Furthermore, the greater the number of cumulative oral antibiotic-treated events over the study period, the steeper the rate of lung function decline ¹⁴ . Consequently, close follow-up post-treatment is warranted to ensure recovery, and patients with repeated exacerbations treated with oral antibiotics may warrant more intensive treatment with IV antibiotics.

Even among patients who receive aggressive treatment with IV antibiotics, PEx outcomes remain suboptimal. Based on CF registry data from both the US and Canada, 25% of individuals who experience exacerbations fail to recover baseline lung function as defined by 90% of baseline within 3 months of treatment ^{2, 15} . Factors independently associated with non-response included female sex; malnourishment; pancreatic insufficiency; persistent infection with Pseudomonas aeruginosa, Burkholderia cepacia complex, or methicillin-resistant Staphylococcus aureus; allergic bronchopulmonary aspergillosis; larger drop in FEV ₁ % predicted at the time of PEx; and longer time from baseline spirometric assessment ² .

It is important to recognize that the proportion of non-responders can vary substantially depending on the definition of “response” used. For example, when a definition of FEV ₁ improvement to 90% of baseline is employed, 25% of exacerbations are classified as non-responders, whereas an improvement to 100% of baseline yields a non-response rate as high as 60% ¹⁶ . Furthermore, it should be noted that up to 25% of patients in the STOP trial had their best lung function at the time of PEx diagnosis and therefore a sizeable proportion of patients will be defined as responders even before treatment has started, leading to underestimation of treatment non-response. Most studies have used the best FEV ₁ in the 3 months following the end of IV antibiotic treatment as the follow-up FEV ₁, since lung function improvement can continue following the completion of IV antibiotics and therefore end-of-treatment values can underestimate rates of response.

In general, the goal of PEx treatment is to improve symptoms and recover lost lung function ³⁴ . Based on the STOP study, physicians identified recovery of lung function as the primary objective of treatment in 53% of exacerbations compared with improvement of symptoms in 43% of exacerbations ⁵ . There is considerable variability as to what constitutes an acceptable threshold for FEV ₁ improvement before antibiotics can be stopped. Based on the STOP study, the mean FEV ₁ improvement was 9% (standard deviation [SD] 10%) predicted at the end of IV antibiotic treatment and 7% (SD 11%) predicted at day 28. Patients with baseline FEV ₁ of more than 50% predicted had a greater increase in FEV ₁ % predicted from admission to day 28 than patients with baseline FEV ₁ less than 50% predicted (10% versus 3%). Interestingly, there was discordance between physician treatment targets in terms of lung function improvement and evaluation of treatment success. Based on the STOP study, 84% of clinicians deemed PEx treatment successful, although only 61% of patients achieved at least 90% of their target FEV ₁ by the end of IV antibiotic therapy ³⁵ .

The treatment endpoint for an individual patient is likely to depend on the primary motivating factor for treatment. In the STOP study, a significant proportion of patients (20%) were admitted for IV antibiotics despite presenting with their best recorded FEV ₁ % predicted in the prior 6 months ⁵ . In these cases, symptoms were the primary driver for treatment and symptom resolution would be the most appropriate endpoint of treatment response. In general, exacerbations are more likely to be defined based on symptoms alone in children compared with adults (in whom drop in lung function may be more likely to occur) ³⁶ . The most widely used daily scoring system for monitoring of respiratory/infectious symptoms during PEx treatment is the Cystic Fibrosis Respiratory Symptom Diary – Chronic Respiratory Infection Symptom Score (CFRSD-CRISS) ³⁷ . Total scores range from 0 to 100, and an 11-point decrease is considered clinically significant ³⁸ . Based on STOP, CFRSD-CRISS decreased by 26.1 (95% confidence interval 23.8–28.3) and 83% of patients achieved a clinically significant improvement ³⁵ . However, it should be noted that the CFRSD-CRISS has limitations in that it has been used and evaluated in a research setting only and has not been accepted by the US Food and Drug Administration as a validated endpoint.

Whereas short-term goals of PEx treatment are to recover lost lung function and improve symptoms, long-term treatment goals generally are to prevent recurrent events and reduce the rate of lung function decline. Interestingly, a recent study found that symptom improvement (that is, CFRSD-CRISS score) in response to PEx treatment is poorly predictive of long-term response in terms of recovery of baseline lung function at 3 months or time to next IV antibiotics; however, immediate FEV ₁ response (>10% relative improvement) was predictive of recovery of baseline FEV ₁ by 3 months ³⁹ . Synthesizing all of this evidence suggests that a composite outcome that includes symptom and lung function improvement might be the most appropriate endpoint in prospective studies examining PEx therapies.

Based on the 2009 Cystic Fibrosis Foundation PEx treatment guidelines, there was insufficient evidence to provide recommendations on a number of decisions related to PEx management, including site of treatment (inpatient versus outpatient), antibiotic duration, and the use of adjunctive therapies (such as systemic steroids) ( Figure 1) ⁴⁰ . Since this publication, observational studies have provided additional insights into these treatment controversies and randomized controlled trials (RCTs) are ongoing or have been recently completed.

There are some circumstances in which admission to hospital for optimal PEx treatment is clearly indicated (such as hypoxia, complications such as pneumothorax, or co-morbidities such as renal failure), but in many situations outpatient treatment with IV antibiotics appears to be a reasonable option. However, in deciding on the optimal treatment setting, an important factor is not just safety and feasibility but also whether one treatment setting is more efficacious than the other. One small randomized, two-factor, mixed-design comparative study involving 17 adults with CF has been performed and showed similar results for most outcome measures for home and hospital treatment ⁴¹ . Owing to selection bias, the results of a larger RCT of the two treatment settings may not be generalizable and will be challenging to interpret because of lack of blinding, attrition bias, high rates of crossover, and lack of standardized resources for home care at various care centers ⁴⁰ .

A small, single-center retrospective study examined 143 PEx events from 50 patients and compared hospital with home IV therapy ⁴² . The baseline characteristics between the two groups were similar; however, the hospitalized group had a greater improvement in lung function with a shorter duration of therapy ⁴² . In a recent observational study of pediatric CF patients from the Epidemiologic Study of CF (ESCF), those with an acute decline in lung function were more likely to recover to within 90% of baseline FEV ₁ when receiving treatment as an inpatient versus outpatient ⁴³ . However, observational studies involving treatment are often confounded by indication bias. In other words, patients are not randomly assigned and therefore the intensity of treatments is not standardized between comparisons; ultimately, patients in hospital might receive more intensive treatments. Another recent study that also used the ESCF employed statistical approaches to control for indication bias and found that PEx with a greater proportion of days treated as an inpatient (versus outpatient) with IV antibiotics was more likely to lead to return of FEV ₁ % predicted to at least 90% of baseline ⁴⁴ . Although this observational study is not definitive (as there was likely residual confounding), the benefits of inpatient PEx treatment are likely due to multiple factors other than the IV antibiotics alone, including improved medication adherence, better nutrition, real-time adjustment of adjunct therapies, increased airway clearance treatments, and increased rest.

Ultimately, the decision regarding treatment site should be made based on careful consideration of patient factors (with a low threshold for inpatient treatment if there are concerns about patient reliability and adherence to airway clearance therapies, sufficient home supports, or more complex IV antibiotic regimens) ⁴⁰ .

For historical reasons, CF exacerbations are typically treated for 14 days. Based on the STOP study, the mean duration of treatment was 15.9 (SD 6.0) days; however, 11% of patients received treatment for 10 days or less and 60% received treatment for more than 14 days with no significant differences in duration for individuals younger than 18 years old versus those older than 18 years old ³⁵ . There have been no randomized studies examining the duration of IV antibiotics to provide evidence-based recommendations. An observational study that used the US CF Foundation Patient Registry found that FEV ₁ % predicted plateaus by day 10 of treatment and duration of treatment did not influence time until next PEx ⁴⁵ . More recent retrospective data on the benefits of prolonging antibiotic treatment are conflicting, as a study using the Toronto CF Database found improved outcomes with longer treatment duration (>14 days) whereas a larger study using data from the ESCF did not find a significant association between treatment duration and rate of recovery of FEV ₁ % predicted to within 90% of baseline ^{16, 44} . However, as in observational studies of treatment setting, interpretation is limited by indication bias, since patient factors influence treatment duration (patients with lower lung function received longer treatment courses) ⁴⁵ . A large multi-center study (STOP2) is under way in the US examining IV antibiotic duration since it was identified as the most important research question by CF physicians and patients/caregivers. A divergent trial design is being used and randomly assigns patients to 10 versus 14 days or 14 versus 21 days of IV antibiotics depending on initial symptom and lung function response by day 7 ⁴⁶ .

A recent study evaluating inpatient PEx treatment practices for pediatric CF patients across the US demonstrated wide variability in the use of adjunctive treatments, including hypertonic saline, azithromycin, and systemic corticosteroids ⁴⁷ . Several adjunct therapies for the treatment of CF PEx have recently been evaluated or are currently under investigation. These adjunct therapies function by either optimizing airway clearance or reducing airway inflammation.

Although nebulized hypertonic saline is well established as a strategy for PEx prevention ⁴⁸ , it has only recently been studied as an adjunct during PEx treatment. In an RCT, Dentice et al. compared PEx outcomes in individuals randomly assigned to nebulized 7% hypertonic saline versus taste-masked control thrice daily ⁴⁹ . The majority of patients in both groups had either never or only intermittently used hypertonic saline prior to enrollment. Although this study did not meet its primary endpoint in terms of reduced length of hospital stay, there was greater improvement in symptoms and higher rates of FEV ₁ recovery in the hypertonic saline group ⁴⁹ . Furthermore, the study provided reassurance that nebulized hypertonic saline was safe to start (or increase) in the context of acute PEx ⁴⁹ .

Another recent study examined doxycycline as an adjunct therapy for CF PEx, acting as a small-molecule inhibitor of matrix metalloproteinase-9 (MMP-9), which has been implicated in CF airway pathophysiology, particularly during PEx ^{50, 51} . This single-center RCT randomly assigned 39 CF patients with PEx requiring inpatient care to either doxycycline 100 mg orally twice daily or placebo for 8 days in addition to standard patient care (IV antibiotics and increased airway clearance) ⁵² . Compared with the placebo group, the doxycycline group had a significant decrease in total and active sputum MMP-9 levels, improved protease-antiprotease imbalance in the airways, greater improvement in FEV ₁ % predicted from admission, and longer time to next PEx ⁵² . Whether doxycycline has direct effects on dysregulated protease activity versus indirect effects due to changes in the microbiome remains unclear, but this single-center study has set the stage for a larger multi-center placebo-controlled RCT ⁵² .

Although systemic corticosteroids are used in up to 20% of CF exacerbations ^{35, 39} , there is limited evidence to support their use ⁴⁰ . Just one small pilot placebo-controlled study involving 24 patients (≥10 years old) examined oral prednisone (2 mg/kg per day up to a maximum of 60 mg divided twice daily) versus placebo for the first 5 days as an adjunct to standard-of-care PEx treatment ⁵³ . Although there was no significant effect on lung function, symptom improvement, or sputum inflammatory markers compared with placebo, the study was underpowered to evaluate treatment effects ⁵³ . A multi-center, randomized, placebo-controlled trial evaluating prednisone—referred to as the Prednisone in CF Pulmonary Exacerbation (PIPE) study—is under way and involves six pediatric and adult CF clinics across Canada. This study will provide more definitive evidence regarding the role of systemic corticosteroids during PEx. Patients receiving IV antibiotic treatment and who have not responded to standard of care alone by day 7 (that is, not recovered at least 90% of their baseline FEV ₁ % predicted) will be randomly assigned to prednisone (2 mg/kg per day up to a maximum of 60 mg divided twice daily) or placebo for 7 days.