G551D, a mutation of the cystic fibrosis transmembrane conductance regulator (CFTR) gene, results in impaired chloride channel function in cystic fibrosis (CF) with multiple end-organ manifestations. The effect of ivacaftor, a CFTR-potentiator, on exercise capacity in CF is unknown. Twenty G551D-CF patients were recruited to a single-centre, double-blind, placebo-controlled, 28-day crossover study of ivacaftor. Variables measured included percentage change from baseline (%Δ) of VO2max (maximal oxygen consumption, primary outcome) during cardiopulmonary exercise testing (CPET), relevant other CPET physiological variables, lung function, body mass index (BMI), sweat chloride and disease-specific health related quality of life (QOL) measures (CFQ-R and Alfred Wellness (AWEscore)). %ΔVO2max was unchanged compared with placebo as was %Δminute ventilation. However, %Δexercise time (mean 7.3, CI 0.5–14,1, P=0.0222) significantly increased as did %ΔFEV1 (11.7%, range 5.3–18.1, P<0·005) and %ΔBMI (1.2%, range 0.1–2.3, P=0·0393) whereas sweat chloride decreased (mean −43.4; range −55.5–18.1 mmol·l−1, P<0·005). Total and activity based domains in both CFQ-R and AWEscore also increased. A positive treatment effect on spirometry, BMI (increased), SCT (decreased) and total and activity based CF-specific QOL measures was expected. However, the lack of discernible improvement in VO2max and VE despite other positive changes including spirometric lung function and exercise time with a 28-day ivacaftor intervention suggests that ventilatory parameters are not the sole driver of change in exercise capacity in this study cohort. Investigation over a more prolonged period may delineate the potential interdependencies of the observed discordances over time. Trial registration number: ClinicalTrials.gov-NCT01937325.

Introduction

Cystic fibrosis (CF) results from mutation of the gene which encodes Cystic Fibrosis Transmembrane conductance Regulator protein (CFTR), a chloride ion channel [1]. The G551D mutation (Class III) affects the ion-gating characteristic of CFTR function. Ivacaftor, a CFTR-potentiating agent is now standard of care for patients with this mutation [2]. Pivotal ivacaftor studies have shown reduced sweat chloride (SCT) and increased FEV1 and weight with treatment up to 48 weeks, improving to maximal response within 4 weeks [2].

Traditionally, parameters tested in clinical trials involving CF patients have measured lung function, body mass index (BMI) and exacerbation rate and these have proven to be useful as a measure of disease severity. They do not provide the integrated, global assessment of cardiovascular, respiratory, muscular and metabolic function that cardiopulmonary exercise testing (CPET) allows [3]. In the clinical setting, CPET has been shown to be a sensitive, repeatable, non-invasive method to assess cardiorespiratory and muscular responses to exercise [4] while some CF standards [5] recommend annual CPET assessment. Exercise capacity is a marker metabolic performance, lung function trajectories and quality of life improvements in CF [6]. Specifically CPET measures oxygen consumption, exercise duration and workload, which are used as prognostic indicators for survival in CF patients [7,8]. This is used to guide individualized medication and therapeutic strategies including exercise programmes and timing of lung transplant referral. Despite this CPET use in clinical trials in CF has been uncommon, however, in the non-CF population some parameters have been shown to be prognostically superior suggesting it may be advantageous in assessing therapeutic responses in CF trials.

Exercise response is impaired in CF [9], while the effect of ivacaftor on exercise is inconclusive [10-12]. CFTR is widely distributed therefore there are several potential mechanisms to explain CF-associated exercise limitation including but not restricted to reduced ventilation and salt and water depletion. With multiple potential cellular and tissue effects in CF patients treated with ivacaftor, CPET represents an ideal method of measuring the physiological response to a therapeutic intervention [13].

The aim of this study was to assess whether a cohort of CF-G551D patients treated with ivacaftor had an improved exercise capacity in a similar time-frame to the previously identified gains in lung function and weight [2]. We chose CPET indices (VO2max, work capacity and duration of exercise) to assess the response to ivacaftor treatment as well as other relevant parameters including SCT, bio-impedance testing (BIA) to calculate body composition and CF-specific quality of life measures.

Methods

Study approval was obtained from the institutional ethics committee at The Alfred hospital (IEC339/13). Following consent, we enrolled patients in a single-centre, double-blind, placebo-controlled, randomized, crossover study of ivacaftor 150 mg BD. Adult patients with at least one copy of the G551D gene mutation were enrolled in the trial. Eligible patients underwent screening tests (day −28) including SCT, routine biochemistry, lung function, BIA and CPET to determine eligibility. Full physical exam, vital signs, anthropometric measurements and laboratory renal, coagulation, full blood count and liver tests were performed at each visit. Testing was repeated at randomization (day 0) and patients were assigned (1:1) to active or placebo treatment. All tests were repeated after treatment period 1 (day 28) and following a 28 day washout period (day 56) [14]; subjects then crossed over to alternate treatments with testing repeated both before and after period 2 (day 84) (Figure 1). Routine CF care was unchanged and patients were encouraged to maintain their baseline activity level; no prescribed exercise programme was included and patients’ exercise activity, and diet was reviewed and reinforced for constancy during study visits.

Consort diagram identifying recruitment and treatment schedules.

Figure 1
Consort diagram identifying recruitment and treatment schedules.

Consort diagram.

Figure 1
Consort diagram identifying recruitment and treatment schedules.

Consort diagram.

Patients were eligible if aged between 16 and 75 years, had a diagnosis of CF confirmed (SCT, genotype and phenotype assessment) and had at least one copy of the G551D mutation. All patients had to be able to perform CPET and complete questionnaires and assessments. Inclusion required a percent predicted FEV1 ≥ 25% at the time of screening. Participants were excluded if deemed unlikely to physically complete a CPET study or had a known adverse reaction to ivacaftor while a positive pregnancy test at screening excluded female participants. Other exclusions comprised strong CYP3A inducers (e.g. rifampicin) or significant liver dysfunction (transaminases above five times the upper limit of normal).

Interventions

Study drug and placebo were supplied by Vertex Pharmaceuticals Inc. A random allocation sequence was generated by a pharmacist and dispensed using sequentially numbered containers. Participants and investigators were blinded until after data were unlocked on completion of all interventions. Adverse events were documented and reported to the institutional ethics committee and Vertex Pharmaceuticals Inc.

CPET was performed using an upright cycle ergometer using a maximal stage 1 incremental protocol (Medgraphics Cardiopulmonary Diagnostics, Med Graphics Corporation, Minnesota, U.S.A.), according to ATS/ACCP guidelines [15]. Work rate was increased by 10–20 watt every minute at a set cadence (60–70) and concluded if not maintained [15]. The Watt increment in load allocated to each patient was based on the modified Godfrey protocol and maintained at the same level for all subsequent studies. Oxygen saturation (SpO2; Nonin 7500 Pulse Oximeter, Nonin Medical INC., Plymouth MN, U.S.A.) was recorded throughout CPET. Metabolic data, ventilatory and cardiac data, were recorded breath by breath and analysed using 20-s averaging [15]. Chest pain and Borg scores (dyspnoea and leg fatigue) were recorded following maximal exercise.

The primary outcome was percent change from baseline for maximal oxygen uptake (%ΔVO2max) recorded at the end of a maximal CPET test based on ATS criteria [15]. Secondary endpoints (Table 2) included additional parameters measured during exercise (minute ventilation, VO2/HR, exercise time, VO2t1/2, Borg scores [8,16]) as well as spirometry [17], weight, BMI and fat free mass (FFM) (Seca mBCA 514™ Body composition analyser, Seca, Hamburg, Germany). Sweat chloride (SCT; Webster Macroduct, Utah, U.S.A.) was measured and assessed for correlation with CPET parameters, lung function and body mass.

The Cystic Fibrosis Questionnaire-Revised (CFQ-R) is a disease-specific health-related quality of life measure for CF patients [18]. The CFQ-R measures subjective assessments in a number of domains which include social functioning, emotional functioning, physical functioning, vitality, health perceptions, respiratory symptoms, treatment burden and role functioning. The Alfred Wellness Score (AWEscore) is a 10-point questionnaire that was developed to readily assess well-being on a visual-analogue scale where 0 = least well and 10 = most well. The wording for anchors of individual questions varied according to the topic to ensure standardized replies. The maximum possible score is 100. The 10 items assessed in the AWEscore are: general health, energy levels; exercise participation daytime, nocturnal cough; sputum; appetite; recommended weight; mood; anxiety level; and sleep (amount and quality).

Statistical analysis

We recruited all available patients into this study; a cross-over study of 20 patients has 80% power to detect a 10% within-patient difference (standard deviation = 15%) in %ΔVO2max between the ivacaftor and placebo arms. Patients’ baseline characteristics were tabulated using mean (SD) or median (interquartile range [IQR]) for continuous variables and n (%) for categorical variables. Paired individual variables were assessed using normal probability distribution plot and non-parametric testing was used when normality was not shown. Period effects were adjusted for using standard ANOVA techniques for cross-over study design. All endpoints were analysed according to the intention-to-treat principle. No adjustment for multiplicity was made but important endpoints and their analysis was pre-specified in a statistical analysis plan (SAP), analysis of other end-points was considered exploratory and hypothesis-generating. Percentages (%) and changes from baseline (Δ) during cross-over were analysed using ANOVA [19], including patients, treatment and period as factors in the model. Corresponding means per treatment arm, 95% confidence intervals (CI), treatment effect (TE) and P-values were reported. The treatment effect is defined as the placebo corrected difference in response (percentage change from baseline) to ivacaftor for any given parameter. When indicated by normality testing ANOVA based on ranks within period was conducted.

Results

Of 340 patients attending The Alfred Hospital Cystic Fibrosis Service, 26 patients were eligible for the study between January and June 2014 based on G551D mutation status. Six patients were excluded due to inability to adhere to study requirements. All patients except two related homozygous individuals were heterozygous for G551D. Demographics at randomization (day 0) are shown in Table 1. One patient fell below FEV1 ≥ 25% at baseline (day 0) but was included based on screening (day −28) results as per pre-defined inclusion criteria. Compliance to blinded study medication (self-reporting, returned blister packs) was above 90% throughout. Three patients decreased study drug dose to avoid potential drug interactions [20]. Study drug (placebo or active treatment) was halved as per pharmaceutical guidelines, investigators remained blinded. A total of 100 CPET studies were performed with 99 included in data analysis. All patients achieved a maximal CPET as per ATS/ACCP guidelines [15]. Comparisons between the beginning of each treatment period day 0 and day 56 indicated no evidence of crossover effect.

Table 1
Patient baseline characteristics (mean, range)
Baseline demographics n=20  
Age (mean) 32 years (18–65) 
Male 12 
Height (cm, mean) 169 (150–182) 
Weight (kg, mean) 67 (49–121) 
BMI mean 25.8 (18–36.4) 
Sweat chloride (mmol/l, mean) 98.05 (56–113) 
Smoker 
Oxygen Supplementation 
Exocrine Pancreatic Insufficiency 18 
CF Diabetes Mellitus 
CF Liver Disease 
Fat Mass (%) 23.0 (4–42) 
Microbiology  
Pseudomonas colonization 19 
MSSA colonisation 
MRSA colonisation 
M abscessus colonisation 
Genotype  
G551D homozygote 
Other mutation DF508 12 
Other mutations G524X (2), V520F (1), Unknown (3) 
Lung function  
FEV1 % predicted mean 54 (23–110) 
FVC % predicted mean 71 (40–96) 
Maximal exercise indices  
VO2max (l·min−11.78 (1.14–3.28) 
VO2max (ml·min−1·kg−126.75 (16.5–34.7) 
Time to VO2max (min) 10.02 (07.40–12.57) 
Work (watts) 132 (80–195) 
VE/VCO2 31 (22–36) 
HR at VO2max 157 (106–183) 
VO2/HR 9.7 (6–16) 
Baseline demographics n=20  
Age (mean) 32 years (18–65) 
Male 12 
Height (cm, mean) 169 (150–182) 
Weight (kg, mean) 67 (49–121) 
BMI mean 25.8 (18–36.4) 
Sweat chloride (mmol/l, mean) 98.05 (56–113) 
Smoker 
Oxygen Supplementation 
Exocrine Pancreatic Insufficiency 18 
CF Diabetes Mellitus 
CF Liver Disease 
Fat Mass (%) 23.0 (4–42) 
Microbiology  
Pseudomonas colonization 19 
MSSA colonisation 
MRSA colonisation 
M abscessus colonisation 
Genotype  
G551D homozygote 
Other mutation DF508 12 
Other mutations G524X (2), V520F (1), Unknown (3) 
Lung function  
FEV1 % predicted mean 54 (23–110) 
FVC % predicted mean 71 (40–96) 
Maximal exercise indices  
VO2max (l·min−11.78 (1.14–3.28) 
VO2max (ml·min−1·kg−126.75 (16.5–34.7) 
Time to VO2max (min) 10.02 (07.40–12.57) 
Work (watts) 132 (80–195) 
VE/VCO2 31 (22–36) 
HR at VO2max 157 (106–183) 
VO2/HR 9.7 (6–16) 

Endpoints

When compared with placebo there was no significant change in primary endpoint %ΔVO2max (TE = −0.8%, 95% CI −6.8–5.3, P=0.79) or workload (TE = 2.6%, CI −3.3–8.5, P=0·37). Additionally, %Δ minute ventilation (VE; mean 4.2, CI −4.0–12.4, P=0.29) was unchanged. However, %Δ exercise time (mean 7.3, CI 0.5–14,1, P=0.022) was significantly increased as were %ΔFEV1 (11.7%, range 5.3–18.1, P<0·005) and %ΔBMI (1.2%, range 0.1–2.3, P=0·039). All changes from baseline at completion are shown by treatment in Table 2. A surrogate marker of intravascular volume and cardiac function (VO2/HR) showed no significant TE, while recovery parameter VO2t1/2, a marker of prolonged oxygen kinetics [21], remained unchanged (P=0.71) as did Borg scores for legs (P=0.64) and dyspnoea (P=0.32) at the end of exercise suggesting consistent subjective exertion.

Table 2
Treatment effect, change from baseline (mean, 95% CI)
EndpointPlacebo % (95% CI)Ivacaftor % (95% CI)Treatment effect % (95% CI)P-value
% Δ exercise time 1.0 (−3.7–5.7) 8.3 (3.4–13.2) 7.3 (0.5–14.1) 0.0222* 
Δmean (s, range) 5 (−20–40) 43 (−20–320)   
% Δ VO2 time  1.5 (−2.3–5.4) 6.0 (2.0–10.1) 4.5 (−1.1–10.1) 0.1102 
% Δ work  1.7 (−2.3–5.8) 4.3 (0.1–8.6) 2.6 (−33–8.5) 0.3648 
Oxygen consumption     
% Δ VO2 5.6 (1.1–10.0) 5.1 (0.5–9.8) −0.4 (−6.8–6.0) 0.8940 
% Δ VO2max 5.0 (0.8–9.1) 4.2 (−0.2–8.6) −0.8 (−6.8–5.3) 0.7910 
ΔMean (range) ml.Kg−1.min−1 1.98 (−1.7–10.6) 1.97 (−2.1–8.0)   
% Δ Oxygen saturation −0.1 (−0.9–0.6) 0.9 (0.0–1.7) 1.0 (−0.1–2.1) 0.0750 
Ventilation     
% Δ RR 4.4 (−1.1–9.9) 1.8 (−4.0–7.5) −2.6 (−10.6–5.3) 0.4949 
% Δ VE 6.1 (0.5–11.8) 10.3 (4.4–16.3) 4.2 (−4.0–12.4) 0.2940 
Cardiac Response     
% Δ HR  1.2 (−1.6–3.9) 3.5 (0.7–6.2) 2.3 (−1.6–6.2) 0.2360 
% Δ VO2/HR 5.0 (−0.6–10.5) 3.1 (−2.8–8.9) −1.9 (-10.0–6.2) 0.6256 
Ventilation/Perfusion     
% Δ VE/VCO2 1.3 (−1.5–4.0) 2.0 (−0.9–4.9) 0.7 (−3.3–4.7) 0.7034 
% Δ Borg score legs 9.6 (−4.9–24.1) 14.3 (−0.3–28.8) 4.7 (−15.9–25.2) 0.6389 
% Δ Borg score dyspnoea 0.9 (−7.9–9.7) 7.0 (−1.8–15.7) 6.1 (−6.3–18.5) 0.3169 
Recovery     
% ΔVO2t1/2 7.0 (−8.0–22.1) 4.0 (−12.7–20.6) -3.0 (−25.5–19.4) 0.7784 
Lung Function     
% Δ FEV1 % Pd 0.4 (−4.3–5.1) 14.1 (9.4–18.8) 13.7 (7.0–20.3) 0.0004 
% Δ FVC % Pd 2·4 (−3·0–7·7) 6.0 (0.7–11.4) 3.7 (−3.9–11.3) 0.3242 
Sweat and Weight     
Δ Sweat chloride (mmol/l) 0.5 (−8.0–9.0) −42.9 (−51.4–34.4) −43.4 (−55.5–31.3) <.0001 
% Δ BMI 0.7 (−0.2–1.5) 1.9 (1.1–2.7) 1.2 (0.1–2.3) 0.0393 
% Δ FFM 0.7 (−0.2–1.6) 1.9 (1.0–2.8) 1.2 (−0.1–2.4) 0.0725 
EndpointPlacebo % (95% CI)Ivacaftor % (95% CI)Treatment effect % (95% CI)P-value
% Δ exercise time 1.0 (−3.7–5.7) 8.3 (3.4–13.2) 7.3 (0.5–14.1) 0.0222* 
Δmean (s, range) 5 (−20–40) 43 (−20–320)   
% Δ VO2 time  1.5 (−2.3–5.4) 6.0 (2.0–10.1) 4.5 (−1.1–10.1) 0.1102 
% Δ work  1.7 (−2.3–5.8) 4.3 (0.1–8.6) 2.6 (−33–8.5) 0.3648 
Oxygen consumption     
% Δ VO2 5.6 (1.1–10.0) 5.1 (0.5–9.8) −0.4 (−6.8–6.0) 0.8940 
% Δ VO2max 5.0 (0.8–9.1) 4.2 (−0.2–8.6) −0.8 (−6.8–5.3) 0.7910 
ΔMean (range) ml.Kg−1.min−1 1.98 (−1.7–10.6) 1.97 (−2.1–8.0)   
% Δ Oxygen saturation −0.1 (−0.9–0.6) 0.9 (0.0–1.7) 1.0 (−0.1–2.1) 0.0750 
Ventilation     
% Δ RR 4.4 (−1.1–9.9) 1.8 (−4.0–7.5) −2.6 (−10.6–5.3) 0.4949 
% Δ VE 6.1 (0.5–11.8) 10.3 (4.4–16.3) 4.2 (−4.0–12.4) 0.2940 
Cardiac Response     
% Δ HR  1.2 (−1.6–3.9) 3.5 (0.7–6.2) 2.3 (−1.6–6.2) 0.2360 
% Δ VO2/HR 5.0 (−0.6–10.5) 3.1 (−2.8–8.9) −1.9 (-10.0–6.2) 0.6256 
Ventilation/Perfusion     
% Δ VE/VCO2 1.3 (−1.5–4.0) 2.0 (−0.9–4.9) 0.7 (−3.3–4.7) 0.7034 
% Δ Borg score legs 9.6 (−4.9–24.1) 14.3 (−0.3–28.8) 4.7 (−15.9–25.2) 0.6389 
% Δ Borg score dyspnoea 0.9 (−7.9–9.7) 7.0 (−1.8–15.7) 6.1 (−6.3–18.5) 0.3169 
Recovery     
% ΔVO2t1/2 7.0 (−8.0–22.1) 4.0 (−12.7–20.6) -3.0 (−25.5–19.4) 0.7784 
Lung Function     
% Δ FEV1 % Pd 0.4 (−4.3–5.1) 14.1 (9.4–18.8) 13.7 (7.0–20.3) 0.0004 
% Δ FVC % Pd 2·4 (−3·0–7·7) 6.0 (0.7–11.4) 3.7 (−3.9–11.3) 0.3242 
Sweat and Weight     
Δ Sweat chloride (mmol/l) 0.5 (−8.0–9.0) −42.9 (−51.4–34.4) −43.4 (−55.5–31.3) <.0001 
% Δ BMI 0.7 (−0.2–1.5) 1.9 (1.1–2.7) 1.2 (0.1–2.3) 0.0393 
% Δ FFM 0.7 (−0.2–1.6) 1.9 (1.0–2.8) 1.2 (−0.1–2.4) 0.0725 

*P-value calculated via ANOVA based on ranks (per period).

Included in SAP.

More particularly, the statistically significant TE in lung function observed in the %∆ FEV1 [TE = 11·7% (P<0·001; 95% CI 5.3–18.1), being attributable to a change of 1.3% (95% CI −3.2–5.9) for placebo versus 13.7% (95% CI 8.5–17.6) for ivacaftor] was not associated with a significant TE in %∆ FVC [TE = 3.7% (95% CI −6.7–14.0, P=0·46)]. Weight increased with %ΔBMI TE of 1.2% (95% CI 0.1–2.3, P<0.039) and although FFM also increased on ivacaftor versus placebo (0.7 versus 1.9; TE 12% (95% CI −0.1–2.4) this was not statistically significant (P=0.072). In an exploratory correlation matrix incorporating the key variables of VO2max, FEV1, SCT and BMI, there was no significant correlation between relative changes in %ΔVO2max and ΔFEV1 (r=0.268; P=0.27), SCT (r=0.233; P=0.32) or BMI (r=0.0418; P=0.86). Nor was a correlation seen between ΔSCT and %ΔFEV1 (r=0.233; P=0.320), %ΔBMI (r=−0.0892; P=0.708) or VO2max (r=−0.0456; P=0.850).

The Alfred wellness score (AWEscore) and the CFQR, are outlined with a full list of their component domains in Table 3. Significant areas of change in the AWEscore included parameters for energy (TE = 1.5, CI 0.3–2.6; P=0.019), exercise (TE = 1.6, CI 0.5–2.6; P=0.006), and general health (TE = 1.0, CI 0.0–2.1; P=0.05). The CFQR findings were similar with significant changes seen in the domains encompassing exercise (TE = 9.7, CI -24.2-43.5; P=0.02), health (TE = 13.5, CI −22.7–49.7; P=0.004), physical (TE = 15.0, CI −17.6–47.6; P<0.001), vitality (TE = 17.4, CI −20.8–55.5; P<0.001), digestion (TE = 8.6, CI −20.3–37.5; P=0.02) and eating (TE = 5.8, CI −15.5–27.1, P=0.03).

Table 3
AWEscore and CFQR domain variables
EndpointBaselinePlacebo (Δ from baseline)Ivacaftor (Δ from baseline)Treatment effectP-value
Alfred wellness score      
Anxiety 7.2 (3.2–11.6). −0.7 (−4.2–2.8) 0.4 (−2.1−3.0) 1.1 (−2.9−5.2) 0.36 
Appetite 8.2 (4.2–12.2) −0.75 (−3.8–2.4) 0.3 (−3.9−4.5) 1.0 (−2.7−4.8) 0.11 
Coughing 5.4 (1.3–9.5) −0.2 (−4.5–4.1) 1.2 (−4.1−6.4) 1.4 (−4.7−7.5) 0.33 
Energy 5.8 (2.2–9.4) −0.5 (−3.7–2.7) 1.1 (−3.3−5.6) 1.6 (3.0−6.3) 0.02 
Exercise 6.0 (1.9–10.1) −1.1 (−4.3–2.0) 0.2 (−4.2−4.5) 1.3 (−2.8−5.5) 0.01 
General health 6.0 (3.0–8.9) 0.8 (−5.2–3.6) 1.3 (−0.9−3.5) 2.1 (−3.2−7.4) 0.05 
Mood 7.4 (4.3–10.5) −0.6 (−3.6–2.4) 0.5 (−2.9−3.9) 1.1 (−2.2−4.5) 0.21 
Sleep 6.2 (2.7–9.7) −1.0 (−5.6–3.6) 0.7 (−3.8−5.3) 1.7 (−4.6−8.1) 0.44 
Sputum 6.0 (2.4–9.5) 0.4 (−4.0–3.2) 1.0 (−3.9−5.9) 1.4 (−3.8−6.6) 0.31 
Weight score 6.7 (1.7–11.7) 0.0 (−2.4–2.4) 0.2 (−4.0−4.5) 0.3 (−4.0−4.6) 0.62 
Total 64.7 (39.6–89.8) −6.1 (−26.3–14.0) 7.1 (−16.8–31.1) 13.3 (−19.0−45.6) 0.02 
CFQR      
Body 75.8 (41.0–110.7) −0.8 (−28.3–26.7) −0.6 (−27.6–26.4) 0.2 (−25.8−26.2) 0.95 
Digestion 84.0 (49.9–118.1) −5.8 (−32.9–21.3) 2.9 (−17.7–23.6) 8.6 (−20.3−37.5) 0.02 
Eating 81.5 (28.2–134.7) 5.9 (−41.1–26.7) 11.7 (−34.6–57.9) 5.8 (−15.5−27.1) 0.03 
Emotion 77.1 (47.2–107.0) 1.3 (−23.4–26.0) 6.5 (−13.4–26.3) 5.2 (−20.3−30.6) 0.09 
Exercise (social activity) 62.1 (25.3–98.8) −4.5 (−26.8–17.8) 5.2 (−21.2–31.5) 9.7 (−24.2−43.5) 0.02 
Health 52.0 (14.0–90.0) −1.5 (−37.0–34.1) 12.1 (−20.7–44.8) 13.5 (−22.7−49.7) 0.004 
Physical 56.0 (14.4–97.5) −3.4 (−30.0–23.3) 11.7 (−15.5–38.8) 15.0 (−17.6−47.6) 0.0006 
Respiratory 54.7 (25.7–83.6) −6.1 (−41.0–28.8) 16.1 (−29.9–62.0) 22.2 (−26.3−70.6) 0.0006 
Role 71.9 (30.8–112.9) 2.2 (−35.4–39.7) 4.3 (−40.6–49.1) 2.1 (−41.3−45.5) 0.67 
Vitality 48.7 (26.0–71.4) −4.7 (−30.8–21.4) 12.7 (−15.3–40.6) 17.4 (−20.8−55.5) 0.0007 
Weight score 72.4 (8.4–136.4) 10.2 (−20.8–41.1) 10.2 (−27.8–48.1) 0.0 (−38.5−38.5) 0.99 
Treatment 60.3 (27.5–93.1) −0.7 (−39.8–34.1) 0.7 (−32.0–33.3) 1.3 (−40.2−42.8) 0.07 
Total 788.9 (527.7–1050.0) −8.4(−215.6–198.8) 88.5(−143.3–320.3) 96.9(−155.3–349.1) 0.003 
EndpointBaselinePlacebo (Δ from baseline)Ivacaftor (Δ from baseline)Treatment effectP-value
Alfred wellness score      
Anxiety 7.2 (3.2–11.6). −0.7 (−4.2–2.8) 0.4 (−2.1−3.0) 1.1 (−2.9−5.2) 0.36 
Appetite 8.2 (4.2–12.2) −0.75 (−3.8–2.4) 0.3 (−3.9−4.5) 1.0 (−2.7−4.8) 0.11 
Coughing 5.4 (1.3–9.5) −0.2 (−4.5–4.1) 1.2 (−4.1−6.4) 1.4 (−4.7−7.5) 0.33 
Energy 5.8 (2.2–9.4) −0.5 (−3.7–2.7) 1.1 (−3.3−5.6) 1.6 (3.0−6.3) 0.02 
Exercise 6.0 (1.9–10.1) −1.1 (−4.3–2.0) 0.2 (−4.2−4.5) 1.3 (−2.8−5.5) 0.01 
General health 6.0 (3.0–8.9) 0.8 (−5.2–3.6) 1.3 (−0.9−3.5) 2.1 (−3.2−7.4) 0.05 
Mood 7.4 (4.3–10.5) −0.6 (−3.6–2.4) 0.5 (−2.9−3.9) 1.1 (−2.2−4.5) 0.21 
Sleep 6.2 (2.7–9.7) −1.0 (−5.6–3.6) 0.7 (−3.8−5.3) 1.7 (−4.6−8.1) 0.44 
Sputum 6.0 (2.4–9.5) 0.4 (−4.0–3.2) 1.0 (−3.9−5.9) 1.4 (−3.8−6.6) 0.31 
Weight score 6.7 (1.7–11.7) 0.0 (−2.4–2.4) 0.2 (−4.0−4.5) 0.3 (−4.0−4.6) 0.62 
Total 64.7 (39.6–89.8) −6.1 (−26.3–14.0) 7.1 (−16.8–31.1) 13.3 (−19.0−45.6) 0.02 
CFQR      
Body 75.8 (41.0–110.7) −0.8 (−28.3–26.7) −0.6 (−27.6–26.4) 0.2 (−25.8−26.2) 0.95 
Digestion 84.0 (49.9–118.1) −5.8 (−32.9–21.3) 2.9 (−17.7–23.6) 8.6 (−20.3−37.5) 0.02 
Eating 81.5 (28.2–134.7) 5.9 (−41.1–26.7) 11.7 (−34.6–57.9) 5.8 (−15.5−27.1) 0.03 
Emotion 77.1 (47.2–107.0) 1.3 (−23.4–26.0) 6.5 (−13.4–26.3) 5.2 (−20.3−30.6) 0.09 
Exercise (social activity) 62.1 (25.3–98.8) −4.5 (−26.8–17.8) 5.2 (−21.2–31.5) 9.7 (−24.2−43.5) 0.02 
Health 52.0 (14.0–90.0) −1.5 (−37.0–34.1) 12.1 (−20.7–44.8) 13.5 (−22.7−49.7) 0.004 
Physical 56.0 (14.4–97.5) −3.4 (−30.0–23.3) 11.7 (−15.5–38.8) 15.0 (−17.6−47.6) 0.0006 
Respiratory 54.7 (25.7–83.6) −6.1 (−41.0–28.8) 16.1 (−29.9–62.0) 22.2 (−26.3−70.6) 0.0006 
Role 71.9 (30.8–112.9) 2.2 (−35.4–39.7) 4.3 (−40.6–49.1) 2.1 (−41.3−45.5) 0.67 
Vitality 48.7 (26.0–71.4) −4.7 (−30.8–21.4) 12.7 (−15.3–40.6) 17.4 (−20.8−55.5) 0.0007 
Weight score 72.4 (8.4–136.4) 10.2 (−20.8–41.1) 10.2 (−27.8–48.1) 0.0 (−38.5−38.5) 0.99 
Treatment 60.3 (27.5–93.1) −0.7 (−39.8–34.1) 0.7 (−32.0–33.3) 1.3 (−40.2−42.8) 0.07 
Total 788.9 (527.7–1050.0) −8.4(−215.6–198.8) 88.5(−143.3–320.3) 96.9(−155.3–349.1) 0.003 

Adverse events

Ivacaftor was well tolerated and no participants ceased treatment during the study. Adverse events included abdominal discomfort (n=2, placebo = 1), elevated creatinine kinase (n=1, screening period), pre-syncope (n=1), joint pain (n=1), diarrhoea (n=1, placebo) and altered liver function tests (n=1, extension) requiring a 7-day suspension of treatment. All 100 CPETs were successfully completed; however one data set was lost due to software malfunction. Serious adverse events included five hospital admissions for pulmonary exacerbations (screening period (n=2), placebo period (n=1), active period (n=2)); the average length of stay was 13.3 days [7-17].

Discussion

CFTR potentiation with ivacaftor is effective in improving lung function, SCT and weight; however the effect on exercise capacity is less well known [2]. Our placebo-controlled, cross-over study showed an improvement in FEV1, SCT and weight and a statistical increase in exercise time, an indicator of fitness [16], although no significant gain in VO2max was observed. Our results were nevertheless in agreement with earlier trials that revealed an ivacaftor treatment effect on spirometric lung function, SCT and body weight within a 28-day treatment period [2]. The lack of discernible improvement in VO2max and minute ventilation despite other positive changes including spirometric lung function and exercise time with a 28 day ivacaftor intervention suggests that ventilatory parameters are not the sole driver of change in exercise capacity in this study cohort. In addition, our results indicate that the significant %ΔFEV1 does not correlate with the (not significant) %ΔVO2max following 28-days of active treatment. This finding is still compatible with previous data where absolute FEV1 at baseline correlated closely with absolute VO2peak (r=0.71) [22]. At a single time-point FEV1 may correlate closely with the absolute VO2max, however, as we have shown the effect of ivacaftor therapy which improves baseline static FEV1 measurement is not mirrored by a commensurate rise in VO2max response.

As exercise parameters did not mirror the effect of ivacaftor on FEV1, SCT and BMI, we questioned whether exercise limitation in CF was solely a factor of ventilatory function as previously described [22] or if non-respiratory factors such as cardiovascular and muscular systems play a greater role than previously suspected [23]. Comparing respiratory parameters, it is noticeable that despite FEV1 improvements there was no significant rise in minute ventilation during exercise, nor was there a statistically significant increase in ventilatory efficiency based on VE/VCO2.

Additional indicators implying a limited impact of increased FEV1 were: (i) respiratory rate and minute ventilation at peak exercise were unchanged despite a maximal test and (ii) Borg scores were unchanged between tests, suggesting subjective equivalence of test effort to baseline. Overall, the FEV1 improvement did not translate into an improved respiratory performance during CPET. The absence of any significant change in oxygen saturation was not surprising as most CF patients tend not to desaturate during exercise except in more severe disease [24]. The improvement in total exercise time but not VO2max suggests patients could exercise longer within an incremental protocol, but without further increase in maximal oxygen consumption. The time course to reach maximal effect for sweat chloride and lung function in G551D patients on ivacaftor is relatively short, reaching near-maximal values by day 15. The change in body mass is a slower process, reaching near-maximal values only at week 16 [2]. Corrected oxygen uptake (ml/kg/min) could potentially increase if fat-free mass rises (with increased muscle metabolism) or fall if fat mass rises disproportionately. Further investigation is required to determine the effect of the mass increase seen in earlier studies.

In normal subjects VO2 is dependent not only on ventilation, but also circulation and muscle end-organ consumption [25]. In CF, reduced VO2 has been attributed to impaired delivery or peripheral utilization of oxygen [9]. Volume contraction typical of CF [26], may compromise maximal exercise capacity because of reduced cardiac output. The response to ivacaftor might have been expected therefore to show a normovolemic exercise response with an improved cardiac output and stroke volume. The VO2/HR change (surrogate stroke volume measure) was not statistically significant despite the large improvement in SCT, suggesting that changes in SCT as a measure of skin chloride channel function may not necessarily correlate with changes in CFTR channel function in other body systems in response to gene-specific treatment.

CFTR is widely distributed permitting gene-potentiation medications to act beyond the respiratory system. Potential non-respiratory exercise effects might occur through improved cardiomyocyte function [27], mitochondrial oxygen uptake and ATP production independent of chloride channel activity [28]. Bassett et al. [29] showed low-intensity exercise can increase mitochondrial enzyme activity 2.2-fold, with minimal effect on VO2. The implication being that early change in mitochondrial activity may not be measurable via oxygen consumption, however, we saw an increase in exercise time which may indicate an early improvement in endurance. These findings are corroborated by the changes seen in the wellness scores which highlight strong improvements in the domains associated with energy, exercise and vitality. To detect changes in VO2max it may require a greater duration of ivacaftor treatment than it did for respiratory, weight and sweat changes that occurred within 28 days. Alternatively, there may be limited response in exercise to gene potentiation treatment. Finally, although recovery post-exercise is a marker of fitness, we did not see a statistical change in the recovery kinetics, that remained unchanged (%ΔVO2t1/2).

The primary limitation for our study is the relatively small sample size. However, given the limited population eligible for this study (4% of CF patients and the logistic requirement to be treatment naive), our sample size was arguably relatively large for a single centre. Additionally, a single-centre approach avoided potential variability in CPET technique between different sites, emphasising the benefits of the placebo-controlled, cross-over design used for the present study. We acknowledge that daily activity was not measured. However, patients were requested to maintain baseline activity levels throughout the study. Although a learning effect cannot be completely excluded, this was minimized by including a baseline CPET (day -28) performed prior to randomization CPETs (day 0) to familiarize patients with testing. Previous studies have shown a minimal learning effect with CPET in CF patients [30].

The present study found that following 28-day ivacaftor gene potentiation treatment there was a significant response in lung function, weight, sweat chloride and exercise time but no evidence of an augmentation in oxygen consumption. Interestingly there was no significant correlation between %ΔSCT and any of %ΔFEV1, %ΔBMI or %ΔVO2max. Given previous evidence showing FEV1 is linked to baseline VO2max, these findings suggest that treatment-specific exercise improvement in CF patients is more complex than improvement in lung function alone. The improvement in exercise time, as well as subjective vitality and mood domains identifies that further evidence linking the differential effects of CFTR on cell, tissue and organ function under steady state and treatment conditions is required. Our results suggest that CF-related skeletal and cardiac muscle dysfunction may be more important than previously appreciated, particularly if CFTR related dysfunction includes changes in cellular energetics and mitochondrial regulation.

In conclusion, in ivacaftor treated CF-G551D patients, positive changes in SCT, BMI and spirometry are not associated with positive changes in VO2max. There was nonetheless an increase in maximal exercise time and improvements in both total and activity-based CF-specific quality of life measures. As well as CPET being clearly useful in measuring global physiological responses that are beyond standard measures of lung function; it also has the potential to provide insights into CF pathobiologies that differentially respond to specific treatment interventions and therefore should be further utilized to more broadly interrogate the global effects of novel gene-potentiator therapies.

Clinical Perspectives

  • Currently there are published studies showing 28-day response in spirometry, sweat testing and BMI in response to ivacaftor treatment in G551D-cystic fibrosis; however, there are no published placebo-controlled studies on the effect that these medications have on exercise in cystic fibrosis (CF).

  • Our study showed a significant treatment effect on ventilatory and quality of life (QOL) parameters over a short 28-day period was not mirrored by a similar VO2max response; however, other parameters indicated improved exercise endurance.

  • In addition to respiratory function other elements of the exercise cascade, such as muscular and mitochondrial function, play a greater role in CF exercise capacity than previously thought, therefore, future studies should include CPET in addition to lung function to fully assess physiological response to therapeutic intervention.

The authors wish to thanks Messrs Matt McGee, Brigitte Borg, Mary Fantidis, Felicity Finlayson and Anthony Talbot, Bill O’Shea for their valuable assistance.

Competing Interests

Dr Wilson and Dr Edgeworth have received speaking consultation fees from Vertex Pharmaceuticals Inc. Dr Wilson, Dr Keating, Dr Kotsimbos and Dr Ellis have a consultancy agreement with Vertex Pharmaceuticals inc. Other authors have no competing interests.

Funding

This study was supported through an investigator initiated grant from Vertex Pharmaceuticals Inc [grant number IIS-2013-103083]; The Alfred Foundation and Mmes Balson and Berry.

Author Contribution

J.W. is the guarantor for this article and had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis, including and especially any adverse effects. D.E., D.K., T.K. and J.W. designed the study. S.H. performed the statistical analysis; D.E., D.K., M.E., E.W., D.C., B.B., A.T., T.K. and J.W. oversaw the study execution, data analysis, interpretation, writing of the manuscript. Study approval was obtained from the ethics committee at the Alfred hospital.

Ethics approval number is: 339/13

Abbreviations

     
  • ANOVA

    analysis of variance

  •  
  • BMI

    body mass index

  •  
  • BIA

    bioimpedence analysis

  •  
  • CFTR

    cystic fibrosis transmembrane conductance regulator protein

  •  
  • CFQ-R

    cystic fibrosis questionnaire-revised

  •  
  • CPET

    cardiopulmonary exercise test

  •  
  • FEV1

    forced expiratory volume in one second

  •  
  • FFM

    fat free mass

  •  
  • FVC

    forced vital capacity

  •  
  • HR

    heart rate

  •  
  • RR

    respiratory rate

  •  
  • SCT

    sweat chloride test

  •  
  • VE

    minute ventilation percent predicted

  •  
  • VE/VO2

    minute ventilation per oxygen consumption

  •  
  • VE/VCO2

    minute ventilation per carbon dioxide production

  •  
  • VO2

    oxygen consumption

  •  
  • VO2max

    maximal oxygen consumption per kilogram

  •  
  • VO2/HR

    oxygen pulse

  •  
  • VO2/t-slope

    oxygen consumption recovery slope

References

References
1
Tsui
L.C.
,
Farrall
M.
and
Donis-Keller
H.
(
1989
)
Report of the committee on the genetic constitution of chromosomes 7 and 8
.
Cytogenet. Cell Genet.
51
,
166
201
,
PubMed PMID: 2676369
2
Ramsey
B.W.
,
Davies
J.
,
McElvaney
N.G.
,
Tullis
E.
,
Bell
S.C.
,
Drevinek
P.
et al
(
2011
)
A CFTR potentiator in patients with cystic fibrosis and the G551D mutation
.
N. Engl. J. Med.
365
,
1663
1672
,
PubMed PMID: 22047557. Pubmed Central PMCID: 3230303
3
Williams
C.A.
,
Saynor
Z.L.
,
Tomlinson
O.W.
and
Barker
A.R.
(
2014
)
Cystic fibrosis and physiological responses to exercise
.
Expert Rev. Respir. Med.
8
,
751
762
,
PubMed PMID: 25395018
4
McKone
E.F.
,
Barry
S.C.
,
FitzGerald
M.X.
and
Gallagher
C.G.
(
1999
)
Reproducibility of maximal exercise ergometer testing in patients with cystic fibrosis
.
Chest
116
,
363
368
,
PubMed PMID: 10453863. Epub 1999/08/24.eng
5
Trust
C.F.
Standards for clinical care of children and adolescents with cystic fibrosis in the UK
,
www.cysticfibrosis.org.uk/media/82070/2011 (accessed on January 29, 2014)
6
Klijn
P.H.
,
Oudshoorn
A.
,
van der Ent
C.K.
,
van der Net
J.
,
Kimpen
J.L.
and
Helders
P.J.
(
2004
)
Effects of anaerobic training in children with cystic fibrosis: a randomized controlled study
.
Chest
125
,
1299
1305
,
PubMed PMID: 15078738
7
Nixon
P.A.
,
Orenstein
D.M.
,
Kelsey
S.F.
and
Doershuk
C.F.
(
1992
)
The prognostic value of exercise testing in patients with cystic fibrosis
.
N. Engl. J. Med.
327
,
1785
1788
,
PubMed PMID: 1435933
8
Moorcroft
A.J.
,
Dodd
M.E.
and
Webb
A.K.
(
1997
)
Exercise testing and prognosis in adult cystic fibrosis
.
Thorax
52
,
291
293
,
PubMed PMID: 9093351. Pubmed Central PMCID: 1758504
9
Hebestreit
H.
,
Hebestreit
A.
,
Trusen
A.
and
Hughson
R.L.
(
2005
)
Oxygen uptake kinetics are slowed in cystic fibrosis
.
Med. Sci. Sports Exerc.
37
,
10
17
,
PubMed PMID: 15632661
10
Quon
B.S.
,
Schaeffer
M.R.
,
Molgat-Seon
Y.
,
Wilkie
S.S.
,
Wilcox
P.G.
and
Guenette
J.A.
(
2015
)
Physiological mechanisms of dyspnea relief following ivacaftor in cystic fibrosis: a case report
.
Respir. Physiol. Neurobiol.
205
,
105
108
,
PubMed PMID: 25447681
11
Saynor
Z.L.
,
Barker
A.R.
,
Oades
P.J.
and
Williams
C.A.
(
2014
)
The effect of ivacaftor in adolescents with cystic fibrosis (G551D mutation): an exercise physiology perspective
.
Pediatr. Phys. Ther.
26
,
454
461
,
PubMed PMID: 25251804
12
Lamhonwah
A.M.
,
Bear
C.E.
,
Huan
L.J.
,
Kim Chiaw
P.
,
Ackerley
C.A.
and
Tein
I.
(
2010
)
Cystic fibrosis transmembrane conductance regulator in human muscle: dysfunction causes abnormal metabolic recovery in exercise
.
Ann. Neurol.
67
,
802
808
,
PubMed PMID: 20517942
13
Weir
E.
,
Burns
P.D.
,
Devenny
A.
,
Young
D.
and
Paton
J.Y.
(
2017
)
Cardiopulmonary exercise testing in children with cystic fibrosis: one centre’s experience
.
Arch. Dis. Child.
102
,
440
444
,
PubMed PMID: 27919886
14
Davies
J.
,
Sheridan
H.
,
Bell
N.
,
Cunningham
S.
,
Davis
S.D.
,
Elborn
J.S.
et al
(
2013
)
Assessment of clinical response to ivacaftor with lung clearance index in cystic fibrosis patients with a G551D-CFTR mutation and preserved spirometry: a randomised controlled trial
.
Lancet Respir. Med.
1
,
630
638
,
PubMed PMID: 24461666
15
American Thoracic Society, American College of Chest Physicians
(
2003
)
ATS/ACCP Statement on cardiopulmonary exercise testing
.
Am. J. Respir. Crit. Care. Med.
167
,
211
277
,
PubMed PMID: 12524257
16
Barakat
H.M.
,
Shahin
Y.
,
Barnes
R.
,
Gohil
R.
,
Souroullas
P.
,
Khan
J.
et al
(
2014
)
Supervised exercise program improves aerobic fitness in patients awaiting abdominal aortic aneurysm repair
.
Ann. Vasc. Surg.
28
,
74
79
,
PubMed PMID: 24332259
17
Miller
M.R.
,
Hankinson
J.
,
Brusasco
V.
,
Burgos
F.
,
Casaburi
R.
,
Coates
A.
et al
(
2005
)
Standardisation of spirometry
.
Eur. Respir. J.
26
,
319
338
,
PubMed PMID: 16055882
18
Quittner
A.L.
,
Buu
A.
,
Messer
M.A.
,
Modi
A.C.
and
Watrous
M.
(
2005
)
Development and validation of The Cystic Fibrosis Questionnaire in the United States: a health-related quality-of-life measure for cystic fibrosis
.
Chest
128
,
2347
2354
,
PubMed PMID: 16236893
19
Senn
S.S.
(
2002
)
Cross-over Trials in Clinical Research
, in:
The AB/BA Design with Normal Data
2nd edn, pp.
364
,
John Wiley & Sons, Ltd
20
Anon
(
2014
)
Kalydeco (ivacaftor) Tablets, Prescribing Information [package insert]
,
Vertex Pharmaceuticals Inc
,
Boston, MA, U.S.A.
21
Pouliou
E.
,
Nanas
S.
,
Papamichalopoulos
A.
,
Kyprianou
T.
,
Perpati
G.
,
Mavrou
I.
et al
(
2001
)
Prolonged oxygen kinetics during early recovery from maximal exercise in adult patients with cystic fibrosis
.
Chest
119
,
1073
1078
,
PubMed PMID: 11296172. Epub 2001/04/11.eng
22
Pastre
J.
,
Prevotat
A.
,
Tardif
C.
,
Langlois
C.
,
Duhamel
A.
and
Wallaert
B.
(
2014
)
Determinants of exercise capacity in cystic fibrosis patients with mild-to-moderate lung disease
.
BMC Pulm. Med.
14
,
74
,
PubMed PMID: 24884656. Pubmed Central PMCID: 4011768
23
Dodd
J.D.
,
Barry
S.C.
and
Gallagher
C.G.
(
2006
)
Respiratory factors do not limit maximal symptom-limited exercise in patients with mild cystic fibrosis lung disease
.
Respir. Physiol. Neurobiol.
152
,
176
185
,
PubMed PMID: 16169290
24
Henke
K.G.
and
Orenstein
D.M.
(
1984
)
Oxygen saturation during exercise in cystic fibrosis
.
Am. Rev. Respir. Dis.
129
,
708
711
,
PubMed PMID: 6426354. Epub 1984/05/01.eng
25
Wasserman
K.
(
1997
)
Diagnosing cardiovascular and lung pathophysiology from exercise gas exchange
.
Chest
112
,
1091
1101
,
PubMed PMID: 9377922
26
Holland
A.E.
,
Wilson
J.W.
,
Kotsimbos
T.C.
and
Naughton
M.T.
(
2003
)
Metabolic alkalosis contributes to acute hypercapnic respiratory failure in adult cystic fibrosis
.
Chest
124
,
490
493
,
PubMed PMID: 12907533. Epub 2003/08/09.eng
27
Jiang
K.
,
Jiao
S.
,
Vitko
M.
,
Darrah
R.
,
Flask
C.A.
,
Hodges
C.A.
et al
(
2015
)
The impact of cystic fibrosis transmembrane regulator disruption on cardiac function and stress response
.
J. Cyst. Fibros. vol 15:34-42.
,
15
,
34
42
PubMed PMID: 26119592
28
Kelly
M.
,
Trudel
S.
,
Brouillard
F.
,
Bouillaud
F.
,
Colas
J.
,
Nguyen-Khoa
T.
et al
(
2010
)
Cystic fibrosis transmembrane regulator inhibitors CFTR(inh)-172 and GlyH-101 target mitochondrial functions, independently of chloride channel inhibition
.
J. Pharmacol. Exp. Ther.
333
,
60
69
,
PubMed PMID: 20051483
29
Bassett
D.R.
Jr
and
Howley
E.T.
(
2000
)
Limiting factors for maximum oxygen uptake and determinants of endurance performance
.
Med. Sci. Sports Exerc.
32
,
70
84
,
PubMed PMID: 10647532
30
Saynor
Z.L.
,
Barker
A.R.
,
Oades
P.J.
and
Williams
C.A.
(
2013
)
Reproducibility of maximal cardiopulmonary exercise testing for young cystic fibrosis patients
.
J. Cyst. Fibros. vol 12:644-650
PubMed PMID: 23726255