Combining Ivacaftor and Intensive Antibiotics Achieves Limited Clearance of Cystic Fibrosis Infections

Abstract

Drugs called CFTR modulators improve the physiologic defect underlying cystic fibrosis (CF) and alleviate many disease manifestations. However, studies to date indicate that chronic lung infections that are responsible for most disease-related mortality generally persist. Here, we investigated whether combining the CFTR modulator ivacaftor with an intensive 3.5-month antibiotic course could clear chronic Pseudomonas aeruginosa or Staphylococcus aureus lung infections in subjects with R117H-CFTR, who are highly ivacaftor-responsive. Ivacaftor alone improved CFTR activity, and lung function and inflammation within 48 h, and reduced P. aeruginosa and S. aureus pathogen density by ∼10-fold within a week. Antibiotics produced an additional ∼10-fold reduction in pathogen density, but this reduction was transient in subjects who remained infected. Only 1/5 P. aeruginosa-infected and 1/7 S. aureus-infected subjects became persistently culture-negative after the combined treatment. Subjects appearing to clear infection did not have particularly favorable baseline lung function or inflammation, pathogen density or antibiotic susceptibility, or bronchiectasis scores on CT scans, but they did have remarkably low sweat chloride values before and after ivacaftor. All persistently P. aeruginosa-positive subjects remained infected by their pretreatment strain, whereas subjects persistently S. aureus-positive frequently lost and gained strains. This work suggests chronic CF infections may resist eradication despite marked and rapid modulator-induced improvements in lung infection and inflammation parameters and aggressive antibiotic treatment. IMPORTANCE Recent work shows that people with CF and chronic lung infections generally remain persistently infected after treatment with drugs that target the CF physiological defect (called CFTR modulators). However, changes produced by modulators could increase antibiotic efficacy. We tested the approach of combining modulators and intensive antibiotics in rapid succession and found that while few subjects cleared their infections, combined treatment appeared most effective in subjects with the highest CFTR activity. These findings highlight challenges that remain to improve the health of people with CF.

Keywords: CFTR modulator; Pseudomonas aeruginosa; Staphylococcus aureus; cystic fibrosis; lung infection.

FIG 1

Study design. Subjects received ivacaftor for 1 week before initiating antibiotics. P. aeruginosa (Pa)-infected subjects received 2 weeks of 2 intravenous (IV) antibiotics simultaneously (including meropenem, tobramycin, colistin, or ceftazidime), followed by 3 months of oral ciprofloxacin and inhaled colistin simultaneously (see Table 1). S. aureus (Sa)-infected subjects received 3.5 months of oral flucloxacillin. We collected the first postantibiotic samples after a 1-month antibiotic-free washout period, and subjects were followed for 31 months in total; the filled circles indicate sample collection times.

FIG 2

Ivacaftor rapidly improved CFTR and lung function, and lung inflammation. (A to D) Gray lines represent individuals, the red line represents the mean (A, B, C) or geometric mean (D), blue triangles represent subject 3, green triangles represent subject 9, ivacaftor treatment is indicated at the top, and the shaded region represents the on-antibiotic period. Subjects 3 and 9 are graphed differently because they became culture negative (see Results). Data with each line labeled by subject number are available in Fig. S2. (A) Sweat chloride. (B) Lung function as measured by forced expiratory volume in 1 second percent predicted (FEV₁pp). (C) Change in FEV₁pp from day 0. (D) Neutrophil elastase per mL sputum. The dashed line represents the limit of detection.

FIG 3

Ivacaftor rapidly improved sputum pathogen density. Gray lines represent individuals, the red line represents the geometric mean, blue triangles represent subject 3, green triangles represent subject 9, ivacaftor treatment is indicated at the top, the shaded region represents the on-antibiotic period, and the dotted line represents the limit of detection (LOD). Data with each line labeled by subject number are available in Fig. S3. (A) CFU per mL of sputum growing on MacConkey agar (presumptive Pseudomonas aeruginosa). (B) CFU per mL of sputum growing on mannitol salt agar (presumptive Staphylococcus aureus). See Fig. S4 for DNA-based measurements of pathogen density.

FIG 4

Ivacaftor and antibiotics reduced pathogen density below detectable levels in two subjects. Black lines represent CFU per mL in spontaneous sputum samples; pluses and minuses indicate culture results from each spontaneous sputum, induced sputum, or cough swab sample; and circles indicate times when induced sputum was attempted but unsuccessful. Ivacaftor is indicated at the top, and the shaded region represents the on-antibiotic period. (A) Staphylococcus aureus (Sa) culture results from subject 3. (B) Pseudomonas aeruginosa (Pa) culture results from subject 9. The asterisk (*) indicates that the culture result was confirmed with species-specific qPCR.

FIG 5

Subjects 3 and 9 do not harbor particularly sensitive bacteria. Each circle indicates the inhibitory concentration (IC) for each tested isolate (n = 35 to 96 per subject), and the error bars represent the median and interquartile range (IQR) calculated from the population of isolates. Results from antibiotics used to treat the respective subjects are reported using black circles with red error bars; results from antibiotics not used to treat the respective subjects are reported using gray circles with black error bars, and the shading indicates the subjects that cleared infection (3 and 9). (A) Staphylococcus aureus (Sa) flucloxacillin ICs. (B to F) Pseudomonas aeruginosa (Pa). (B) Ciprofloxacin ICs. (C) Colistin ICs. (D) Meropenem ICs. (E) Ceftazidime ICs. (F) Tobramycin ICs.

FIG 6

Subjects 3 and 9 clinical characteristics. Gray circles indicate individual subjects; the red line is the average (A to E, K, and L) or geometric mean (F to J), blue triangles indicate subject 3 (cleared S. aureus) and green triangles indicate subject 9 (cleared P. aeruginosa). All data are from baseline (day 0). (A) FEV₁ percent predicted. (B) BMI. (C) Age. (D) Total Brody CT score. (E) Brody CT bronchiectasis subscore. (F) CFU per mL of sputum growing on MacConkey agar (presumptive Pseudomonas aeruginosa). (G) CFU per mL of sputum growing on mannitol salt agar (presumptive Staphylococcus aureus). (H) Neutrophil elastase per mL sputum. The dotted line indicates the limit of detection (LOD). (I) IL-1β per mL sputum. (J) IL-8 per ml sputum. (K) Shannon alpha diversity of sputum.

FIG 7

Subjects 3 and 9 had low baseline and ivacaftor-induced sweat chloride. Black lines represent individuals, the red line represents the mean, blue triangles represent subject 3 (cleared S. aureus), and green triangles represent subject 9 (cleared P. aeruginosa).

FIG 8

The relative abundance of MLST types recovered from subjects’ sputum. Each color represents a unique MLST type and is consistent across subjects; d, day; m, month; Abx, antibiotic period. MLST type abundances were determined using PopMLST (see reference 35) and represent the average relative abundance of alleles from 6 P. aeruginosa (Pa) or 7 S. aureus (Sa) MLST loci inferred to originate from a single strain. Where total relative abundance does not equal 100%, less than one-half the MLST loci exhibited a secondary allele, likely due to sequencing error. Symbols differentiate reasons for missing data (see key). (A) Relative abundance of P. aeruginosa MLST types. (B) Relative abundance of S. aureus MLST types. (C) Percentage of subjects who experienced at least one S. aureus strain gain or loss in 1 year for the control cohort (CF subjects not receiving combined ivacaftor + antibiotic treatment [n = 11]) and the combined treatment cohort (this clinical trial of ivacaftor + antibiotics [n = 7]) (**, P = 0.0022).