Plasma and cellular ivacaftor concentrations in patients with CF

Abstract

Access to CFTR modulators has been gradually increasing for people with cystic fibrosis, the first of which was ivacaftor, a CFTR potentiator which is part of all clinically available modulator treatments. In this study, we hypothesized that the steady-state concentrations in blood and tissue are highly variable in patients taking ivacaftor in a real-world context, which may have impacts on treatment approach. We collected nasal epithelial cells to estimate target site concentrations and blood samples to estimate pharmacokinetic parameters at steady state. We found that patients on ivacaftor monotherapy have variable concentrations well above the maximal effective concentration and may maintain concentrations necessary for clinical benefit even if dosing is reduced. We also are the first to provide detailed target site concentration data over time, which shows that tissue concentrations do not fluctuate significantly and do not correlate with plasma concentrations. These findings show that some patients may have higher-than-expected concentrations and may benefit from tailored dosing to balance clinical response with side effects or adherence needs.

1. Introduction:

For persons with cystic fibrosis (PwCF), mutations in the gene that encodes CFTR (Cystic Fibrosis Transmembrane Conductance Regulator) result in diminished CFTR function in multiple organ systems. For the lung, impaired airway clearance, infection, and inflammation causes morbidity and limits life span. ¹ Since 2012, access to mutation-specific therapy using small molecule compounds has continued to expand for PwCF, revolutionizing the approach to CF care. ^2–14 Ivacaftor is currently a component of all FDA-approved modulator therapies and potentiates CFTR channel opening, restoring proper regulation of the mutant ion channel. Long-term studies of ivacaftor monotherapy have shown robust responses resulting in maintenance of lung function and other measures over several years. ^15–18 Most PwCF on ivacaftor monotherapy have experienced significant improvement in multiple clinical outcome measures, categorizing it as a “highly effective modulator therapy” (HEMT). More recently, the introduction of triple combination therapy (which includes ivacaftor as a key component) in 2019 expanded HEMT to the majority (~90%) of PwCF. ^{12; 19} Some PwCF are unable to achieve such robust responses, while others have intolerable side effects, such as liver injury. These differences may be due to variability in drug exposure and/or differing sensitivities to modulator drugs. In this study, we hypothesized that comprehensive steady-state concentrations may vary among PwCF which could influence response to ivacaftor.

PwCF have differences in absorption, metabolism and transport of drugs compared to other individuals, which are known to cause variability in drug response. ^20–23 Furthermore, some clinical studies of ivacaftor have shown highly variable ivacaftor plasma concentrations. ^24–27 These studies are limited, because some measure concentrations only at a single or few time-points ^{24; 26–28}, after a single dose of ivacaftor ²⁹, or only in healthy volunteers. ²⁵ Only one study has attempted to quantify the target-site (nasal epithelial cell) concentrations, which may differ from the concentration in plasma. ²⁸ Our study aims to 1) provide a comprehensive steady-state pharmacokinetic study mirroring typical clinical use for estimation of the variability of concentrations in a real-world setting and 2) assess steady state target-site concentrations of modulators over time.

Methods:

Human subjects.

Patients were identified through clinic lists and via chart review. Patients were consented under an approved human subjects protocol (UAB #300001194). If enrolled, patients were scheduled for up to three visits for blood and nasal brushing collection. Any PwCF on ivacaftor monotherapy between the ages of 2 and 70 years old were eligible for inclusion. Patients with a history of generalized bleeding disorders, recent (within two weeks) or recurrent epistaxis, history of nasal surgery within the past six months, and/or allergies to cleansing or numbing agents used for the nasal brushings, were excluded. Current medications were also reconciled with patients and medical record during initial screening process. Patients taking medications that are listed as contraindicated or listed as known to cause drug-drug interactions by the manufacturer were also excluded from the study.

Blood and tissue sampling:

Patients were scheduled for three visits. Patients were instructed to take their ivacaftor as per manufacturer’s recommendations (150 mg every 12 hours) for at least five days to ensure steady-state concentrations in the run up to Visit 1. Participants were provided high fat/high calorie CF meal options (reviewed by the clinical research unit registered dietician) at Visit 1, and instructed to ingest the medication with fat-containing food as per the manufacturer’s instructions at all visits. ³⁰ Patients were also instructed to continue taking all prescribed medications (including pancreatic enzymes) as instructed by their healthcare provider, and to inform the research coordinator of any changes during study enrollment. Plasma was collected at time 0 (prior to ingestion of first morning dose and 12 hours after the prior evening dose) and at 1, 2, 4, 5, 6, 8, 10 and 12 hours after ingestion. Nasal epithelial brushing was collected at 0 and 6 hours. During Visit 2 and 3, patients were also recommended to take medications as per manufacturer’s instructions for the 5 days prior to the visit, although prepared meals were not provided. ³⁰ Patients were requested to arrive for plasma and nasal epithelial brushing to collect a 0h (pre-dose) sample, and then another between 4–6 hours after ingestion. Quantitation of ivacaftor and its metabolites, M1-ivacaftor (~1/10 activity of ivacaftor) and M6-ivacaftor (considered inactive), was performed as previously described, using an LC-MS/MS assay validated to standards recommended by the FDA for pharmacokinetic studies. ^{28; 30} Cytology brushes were used to collect cells from the inferior turbinate of patients during each visit as described in methods and as reported previously. ²⁸ Cells were counted and lysed, and total ivacaftor concentration was determined. For comparisons to plasma, data was normalized to estimated cell volume as previously reported. ²⁸

Analysis:

After quantitation, non-compartmental analysis (NCA) methods Phoenix 8.1 (Certara, Princeton, NJ) were used to calculate the AUC₁₂; time to reach maximum concentration in plasma (T_max); maximum concentration in plasma (C_max); half-life (t_1/2), oral clearance (CL/F) and apparent distribution volume (V/F). We collected nasal epithelia at 6h post-dose to determine cellular ivacaftor concentrations. At visits 2 and 3, repeat collection coincided with plasma concentration collections. We used the effective concentration to achieve 90% of the maximal clinical effect on lung function (EC₉₀), 250 ng/mL (EC₈₄ for sweat chloride), as a benchmark for comparison. The primary analysis used regression models assuming linear association. For comparisons across visits, generalized linear mixed models, assuming a non-normal distribution, was used for comparisons.

Clinical data analysis:

Clinical outcome data was collected using the electronic medical record and supplemented by data collected in the local centers’ Cystic Fibrosis Foundation Patient Registry (CFFPR), when available. Records from the research unit regarding participation in clinical studies of ivacaftor were also used to determine start dates of ivacaftor, if patients had previously participated in trials. Baseline outcome measures of body mass index (BMI), percent predicted forced expiratory lung volume in one second (ppFEV1), and percent predicted forced expiratory flow 25–75% (ppFEF 25–75) were determined by examining encounter dates that occurred in the absence of pulmonary exacerbations as noted in the CFFPR. Baseline values were calculated by taking the mean of three non-exacerbation values in the year prior to ivacaftor initiation. In some cases, patient baseline was established with fewer than three values, if three were not available. The absolute change in outcome measures within the first year of treatment (best value in the absence of an exacerbation in the first year of use), and at the time closest to enrollment in this study, was calculated. This approach was selected because all patients had been on ivacaftor for a prolonged period at enrollment, so that available prospective clinical data was not available at the same time-points for every participant. Instead, clinical data was extracted from review of the medical record. Encounters that indicated a pulmonary exacerbation were not included in these absolute change calculations. Several patients did not have post-ivacaftor sweat chloride values; these were collected as part of the study if not otherwise available and if the patient consented to the test. Change from baseline in ppFEV1, sweat chloride, and body mass index were calculated. Exploratory analyses included Pearson correlation or Spearman rank, as needed, between concentration data and clinical outcomes.

Results.

Patient demographics and baseline clinical characteristics.

Sixteen PwCF were recruited to this study, demographics are shown in Table 1. The majority of participants were adults, with a median age of 24.5 years (range 9 – 50), 56% female, and a median 6.1 (range 1.8 – 7.3) years prescribed ivacaftor (including times during clinical trial participation, if known). The majority were CFTR G551D heterozygotes (n=12) with 4 patients having another gating mutation and 10 with F508del as the other allele. Three participants were pancreatic sufficient. In Table 1, their pre- and post-ivacaftor clinical characteristics are described, with baseline ppFEV1 of 80%, BMI of 20.5, and sweat chloride of 98 mmol/L. Post-ivacaftor (change in the first year of use, and the most recent values prior to enrollment in the pharmacokinetic study) responses to ivacaftor in this population was overall similar to prior reports including for long-term outcomes. During Visit 1, one patient opted not to participate due to the time commitment and another had difficult intravenous access (n=1), resulting in 14 of the 16 participants providing sufficient data for full PK parameter determination. Fourteen patients from Visit 1 participated in Visit 2 and 12 participated in Visit 3. One participant participated only in Visit 3, and data from that participant is included only in descriptive data. One participant had ivacaftor dose reduction due to clinical concerns and provided additional samples (not included in primary analysis) for evaluation on a reduced dose (150mg once daily).

Table 1.

Participant demographics and clinical characteristics pre- and post-ivacaftor.

Demographics	All (n =16)
Age, median (range)	24.5 (9 – 50)
Female, n (%)	9 (56)
Years on ivacaftor, median (range)	6.1 (1.8 – 7.3)
Genotype
G551D heterozygote, n (%)	12 (75)
Other gating mutation, n (%)	4 (25)
F508del heterozygote, n (%)	10 (62.5)
Pancreatic sufficient, n (%)	3 (19)
Race, n (%)
White	15 (94)
Black/African American	1 (6)
Clinical Characteristics	Pre-ivacaftor	Post-ivacaftor	Mean Absolute Change	p-value
FEV1%Predicted mean (SD), (n = 15) a	80 (22)
In 1st year, (n = 15)		88 (25)	8 (8.2)	0.0026
Most recent prior to study, (n = 14) a		80 (26)	−0.32(10)	0.9823
FEF25–75%Predicted mean (SD), (n = 15) a	77(42)
In 1st year, (n = 15)		84 (45)	8.8 (17)	0.2433
Most recent prior to study, (n = 14) a		65 (39)	−13 (22)	0.0565
BMI kg/m2, mean (SD), (n = 16)	20.4 (4.4)
In 1st year, (n = 16)		22 (4.2)	1.7 (1.3)	0.0002
Most recent prior to study, (n = 15)a		22 (4.4)	2.3 (3.2)	0.0136
Sweat chloride (mmol/L) mean (n, SD)b	98 (15, 31)	50.8 (12, 23)	−46.7 (11, 29.8)	0.0004

Not all participants had comparable timed data points due to retrospective record review, which accounts for missing data.

^aOne participant was a lung transplant recipient and was excluded from pulmonary analyses. One participant did not have clinically-derived spirometry available recent to the visit.

^bOne participant did not have a baseline sweat chloride available and four did not have a post-treatment value available.

Plasma pharmacokinetic parameters.

PK parameter determination was completed for ivacaftor, M1-ivacaftor and M6-ivacaftor and are described in Table 2. Data from these individuals were compared to the manufacturer’s results from Study 008 Part B and Study 010. ³⁰ In comparison to manufacturer study results, the mean area under the curve over 12 hours (AUC₁₂), a measure of total drug exposure, was found to be approximately 67% greater than in similar trials (19.4 compared to 11.6 ng/mL*hr). ³⁰ Minimum (C_min) and maximum concentrations (C_max) were 959 (compared to 636 ng/mL) and 2385 (compared to 1390 ng/mL). The time to C_max (T_max) was similar at 4.4 hours (compared to 4.0 hours). The half-life of ivacaftor was shorter in our study (7.94 compared to 14.7 hours). In Figure 1, concentrations at steady-state over time of ivacaftor is shown, with dotted lines representing each patient’s concentration-time curve, and the solid line is the median for the group. Concentrations of M1-ivacaftor were somewhat lower than expected based on trial results (AUC₁₂ was 37.9 compared to 48.1 ng/mL*hr; C_max was 4570 compared to 5800 ng/mL), with a lower metabolite:parent ratio of 1.95 (compared to 4.89), suggesting a lower production of M1-ivacaftor metabolite in this study. ³⁰ Concentrations of M6-ivacaftor were somewhat higher than expected, likely reflective of the increased ivacaftor seen in these patients.

Table 2.

Non-compartmental analysis parameter results.

	Ivacaftor (n = 14)			M1-ivacaftor (n = 14)			M6-ivacaftor (n = 14)
	Mean (SD)	Median (min, max)	CV (%)	Mean (SD)	Median (min, max)	CV (%)	Mean (SD)	Median (min, max)	CV (%)
Cmin (ng/mL)	959 (733)	771 (64–2610)	76	1950 (1234)	1760 (20–4930)	63	2396 (1962)	1885 (20.5–6030)	82
Cmax (ng/mL)	2385 (1937)	1690 (1070–8170)	81	4571 (2109)	4440 (1520–9050)	46	4397 (2860)	3525 (706–9060)	65
Cavg (ng/mL)	1618 (1263)	1312 (708–5140)	78	3156 (1562)	2799 (1128–6676)	50	3821 (2367)	3548 (627–7891)	62
AUC12	19 (15)	16 (8.5–62)	78	38 (19)	34 (13–80)	50	46 (28)	42 (7.5–95)	62
Tmax (h)	4.4 (0.94)	4 (2–6)	21	5 (1.2)	5 (4–8)	24	6 (3)	6 (0–10)	50
T1/2 (h)	7.9 (2.8)	7.3 (4.2 −15)	35	9.3 (5.3)	7.8 (3.3–21)	57	11 (6.7)	9.9 (5.2–26)	58
CLss/F (L/h)	10.5 (4.6)	9.5(2.4–18)	44	5.0 (2.7)	4.5 (1.9–11)	54	5.9 (5.5)	3.8 (1.7–21)	94
V/F	120 (60)	132 (23–207)	50	70 (70)	45 (15–291)	100	110 (97)	88 (22–307)	88
MPR	-	-	-	2.2 (0.82)	2.2 (0.95–3.7)	37	2.6 (1.7)	2.3 (0.46–6.4)	66

SD, standard deviation; min, minimum value; max, maximum value; CV, coefficient of variation; T_1/2, half-life; T_max, time to maximum plasma concentration; C_max, maximum plasma concentration; C_min, minimum plasma concentration; AUC₁₂, area under the plasma concentration time curve from time of administration to 12 hours after dosing; CL/F, oral clearance; V/F, apparent distribution volume; MPR, metabolite to parent ratio.

Figure 1. Concentration-time curve for ivacaftor concentrations in plasma.

Concentration-time curves (n = 15) for A. Ivacaftor (shaded portion represents level of EC₉₀), B. M1-Ivacaftor, and C. M6-Ivacaftor are shown (dotted lines). Data is presented on a log scale y-axis. The solid black line are median values for each indicated time point.

Plasma concentrations across visits remain above published EC₉₀.

The results of plasma samples contributed by patients at 0 hours (pre-ingestion) and around the peak (4–6 hours after ingestion) at Visits 1, 2, and 3 are shown in Figure 2. Plasma concentrations were significantly higher (p<0.01) at both 0 and 4–6 hours at Visit 1 than at Visit 2 and 3, with no significant different difference between Visits 2 or 3. As described in methods, patients are provided food limited to high fat/high calorie options and medication administration recommendations for Visit 1, while Visits 2 and 3 have more flexibility. For ivacaftor monotherapy, the EC₉₀ is 250 ng/mL, which is the goal concentration of ivacaftor to maintain. ³⁰ In this sample, all patients achieved concentrations near or above the EC₉₀ throughout the entire dosing interval and all three study visits. In our cohort, over all 3 visits, only 2 participants had any concentrations <250 ng/mL; one at time 0 (0 ng/mL) and 1h (64 ng/mL) for Visit 1 and another at 0 (180 ng/mL) and 4h (205 ng/mL) for Visit 3. The average concentration over the entirety of Visit 1 shows that participants exceeded the EC₉₀ by 3–20 fold. Over Visits 2 and 3, the average concentration was 712 ng/mL, nearly 3 fold higher than the EC₉₀.

Figure 2. Timed concentrations of ivacaftor in plasma over Visits 1, 2, and 3.

Matched samples (n = 7 – 15 depending on time point and Visit) from the same participants at 0h and 4–6h from Visits 1, 2, and 3 were plotted on a log scale to compare variability over time. Shaded region represents level of EC₉₀. Generalized linear mixed model analysis was completed for these repeated measures. *p < 0.01, ns = non-significant.

One patient was re-enrolled after their primary pulmonologist reduced dosing of ivacaftor for clinical reasons, to 150 mg once daily. The participant provided two additional samples (0h and 4–6 h after ingestion) over two different visits. The concentrations decreased to 523 ng/mL (from 2700 ng/mL) and to 1240 ng/mL (from 4610 ng/mL). These concentrations remained well above the EC₉₀ of 250ng/mL, although halving the dose resulted in a decrease in plasma concentration of 3.7–5 fold.

Concentrations of ivacaftor and its metabolites in epithelial cells.

As the target site for ivacaftor, epithelial cell concentrations are useful for understanding airway epithelia exposure. Human nasal epithelial cells were used as a surrogate for the airway epithelia due to their overall similarity ³¹, accessibility, and similar expression of drug metabolism protein expression. ³² The mean concentration (standard deviation) prior to ingestion (0h) was 605 (755), 534 (313), and 865 (552) ng/mL for Visits 1, 2, and 3, respectively; for 4–6 hour concentrations the means were 636 (626), 693 (484), and 1200 (1011) ng/mL. The coefficient of variation ranged from 59 – 125%. While there was considerable intra-individual variability, there were no significant differences between time-points (in contrast to the change over time for plasma concentration) nor across visits (Figure 3), suggesting that the cellular concentrations fluctuate less over time compared with plasma. M1-ivacaftor cellular concentrations were detectable in over 95% of samples (Table 3); M6 was quantifiable in only one sample (267 ng/mL).

Figure 3. Timed concentrations of ivacaftor in cell lysates over Visits 1, 2, and 3. A.

Concentrations across visits at 0h. Sample size listed in Table 3. B. Concentrations across visits at 4–6h. C. AUC₁₂ estimates from cell lysate concentrations. Limited sampling strategies allow estimation of the AUC₁₂ using 0h as the 12h given the q12h dosing interval using the C1+C2 method. All had paired plasma data. Mean total plasma and total cellular AUC were significantly different (paired t-test). D. Data from V1 was used to estimate cellular half-life (samples collected at 0 and 6h). 0h is flipped to 12h given q12h dosing, and of the 12 with detectable ivacaftor for this analysis with both time points available, half-life could be estimated from 7. The remaining 5 could not, because the 12h sample was higher than the 6h sample. *p = 0.0194, ns = non-significant.

Table 3.

Ivacaftor and M1-ivacaftor cellular concentrations

	ivacaftor			M1-ivacaftor
	Mean (SD) ng/mL	Median (range) ng/mL	%CV	Mean (SD) ng/mL	Median (range) ng/mL	%CV
Visit 1: 0 hours n=13	605 (755)	295 (0 −2630)	125	696 (1014)	240 (67–3495	146
6 hours n=13	636 (626)	332 (0 −1896)	98	526 (452)	351 (135–1595)	86
Visit 2: 0 hours n=10	534 (313)	595 (0 – 999)	59	398 (166)	376 (86–662)	42
4–6 hours n=13	693 (484)	694 (40 −1627)	70	542 (355)	432 (79–1299 )	66
Visit 3: 0 hours n=8	865 (552)	816 (0 −1575)	64	760(557)	605 (0–1608)	73
4–6 hours n=8	1200 (1011)	1006 (135–2712)	84	879 (752)	804 (0–2031)	86
t1/2 (hours)a	12.3 (13.2)	7.6 (1.7 –39.7)	107	-	-	-
AUC12 (mgxh/L) n=13	7.8 (6.8)	4.62 (2.84 –26)	87	-	-	-

SD, standard deviation; CV, coefficient of variation; T_1/2, half-life; AUC₁₂, area under the plasma concentration time curve from time of administration to 12 hours after dosing.

^aAs described in the manuscript, 7 paired samples provided data that could be used to estimate cellular half-life. Not all participants consented or tolerated repeated nasal brush biopsy for the data in this table.

In Figure 3C, the AUC₁₂ of cellular concentrations from timed nasal brushings were determined and compared to matched plasma samples. Previously, we found limited evidence of several fold increase in concentrations in cellular ivacaftor in a small number of patients.²⁸ In this study, we did not see evidence of such intracellular accumulation in vivo when comparing total ivacaftor concentrations in the cell versus the paired sample in plasma. The t_1/2 was able to be estimated from 7 participants; the remainder either did not have detectable ivacaftor or had concentrations at 0h that were higher than at 4–6h. There was no significant difference between mean t_1/2 in plasma and cellular samples. For patients whose t_1/2 could not be estimated, the plasma half-life was significantly longer than those for whom the cellular t_1/2 could be estimated. There was no significant association between matched plasma and cellular concentrations.

Exploratory analysis of concentration data with clinical response.

Although this study was not powered to detect associations with clinical outcomes, we performed exploratory analysis for absolute changes in ppFEV1, sweat chloride, and BMI from chart reviewed data to compare to PK parameters (AUC₁₂, peak and trough concentrations). No significant associations were detected, although this result was anticipated since all concentrations were above the EC₉₀. ³⁰ Exploratory analysis of trough plasma concentrations demonstrated the following non-significant relationships: sweat chloride change from baseline (r = −0.33, p = 0.35); largest change in ppFEV1 (r = −0.41, p = 0.13); largest change in BMI (r = 0.32, p = 0.26). Assessment of associations was also completed against weight at baseline, under the hypothesis that a smaller weight would result in higher concentrations. ²⁶ For trough concentration and AUC₁₂, there was no significant association with weight or body surface area, however, no participant was less than 43kg, in contrast to the prior study.

Concomitant medications.

All concomitant medications are shown in Supplemental Table 1. Patients reported the use of 87 unique medications throughout this study. Concomitant medications were reviewed at each visit to ensure that no medications that would necessitate a change in therapy (e.g., azoles, rifampin). The most common medications at Visit 1 were azithromycin (87%); albuterol (73%); pancreatic enzymes (67%), dornase alfa (53%); an inhaled steroid (60%); tobramycin (47%) and insulin glargine (33%). All other medications were being used by 27% or fewer patients, and only one patient was on each of the remainder of the medications. Patients were taking an average of 13.6 different medications at Visit 1 (range 1 – 29).

Discussion.

We describe the concentrations of ivacaftor and its primary metabolites in PwCF during typical clinical use. The addition of this study’s real-world PK data to clinical trial data suggests that PwCF taking ivacaftor monotherapy are unlikely to experience subtherapeutic concentrations. Despite significant inter-individual variability in plasma concentrations, all PwCF achieved concentrations above minimal requirements and most had concentrations several times higher than the EC₉₀ reported by the manufacturer (250ng/mL). This threshold was chosen for comparison in our study because this concentration is the EC₉₀ for lung function response, and the EC₈₄ for sweat chloride, providing a suitable target concentration for plasma for both pulmonary and extra-pulmonary responses. Only four time points from two individuals during Visits 1 and 3 measured at concentrations <250 ng/mL. This represents a small fraction of all samples tested, and shows that for all subjects, the majority of concentrations were well above the EC₉₀, regardless of whether the medication administration and diet were controlled (Visit 1) or not (Visits 2–3). ³⁰ Cellular concentration data provide evidence that target site concentrations may not significantly vary over the hours following ingestion, unlike the concentration variation in plasma.

A key finding of our study revealed that some patients have higher than expected ivacaftor concentrations which may be more than required to elicit a robust response. Such patients could potentially benefit from changing the dose. If a PwCF’s concentration is higher that the EC₉₀, of 250ng/mL, they may have little additional clinical improvement. We observed two PwCF with concentrations considerably higher than that observed in clinical trials. ³⁰ Our previous publication described a limited number of PwCF across several modulator combinations ²⁸, whereas this study was designed to focus on ivacaftor-monotherapy PwCF. The mean total cellular concentration of ivacaftor over all samples regardless of time collected was approximately 745 ng/mL (1.9µM). The concentrations required in vitro to potentiate G551D CFTR and other mutations are in the 100nM range. ^{5; 33; 34} Our studies show that total cellular ivacaftor concentration is well above this threshold for the majority of participants, and only a small fraction, ~4% of cellular concentrations, were lower. These data suggests that some patients may tolerate reduction of their dose while still maintaining target-site concentrations sufficient for maximal effect. While not currently clinically available, concentrations can be directly measured, which could then guide tailored therapy. Additional study for concentration-controlled dosing may be required.

During the study, one PwCF underwent reduced dosing by their primary pulmonologist, and we had the opportunity to repeat limited testing. Despite the halving of their dose to 150mg once daily, the trough concentration remained above the EC₉₀. Some PwCF experience intolerable side effects and need to reduce or halt treatment, such as liver inflammation or early breast development. ^{26; 35} Side effects for triple combination therapy (12 years and older dosing), which is expected to have similar concentrations of ivacaftor, are prevalent, and include a wide range of effects such as elevated liver function tests, mental health changes, weight gain, and pain. ³⁶ Our data suggest that some PwCF may have elevated concentrations, would likely tolerate reduction in dose, and that monitoring of concentrations can help understand dose reduction to manage side effect profiles.

An interesting finding was the difference in concentration profiles when drug administration and diet are controlled. In the days leading up to Visit 1, participants were provided specific advice on medication administration, primarily to ensure that they were at steady-state with dosing intervals that were feasible for the 12 hour collection period (8am and 8pm). They were advised to continue a standard CF diet and provided with a high fat/high calorie diet on the day of the study. Ivacaftor exposure is known to increase 2–4 fold with fat-containing meals ³⁵, which may have contributed to the difference in concentrations at Visit 1. However, the quantity of fat needed to adequately absorb ivacaftor is not known, and the classic high fat CF diet may contribute to excessive weight gain in the context of modulator use. ¹⁵ As dietary recommendations are revised in the CF population, modulator concentrations may change. Studies to assess dietary interventions to promote both optimal health and maintain sufficient modulator absorption are needed. Additionally, the influence of concomitant medications suggested, but not proven, to have an impact on ivacaftor concentrations is unknown but warrants further investigation. ^{26; 29; 37; 38}

This study is limited by small sample size. Comprehensive pharmacokinetic studies such as this are burdensome. The study was ongoing at the time of the onset of the global COVID-19 pandemic, and at the time of approval of triple combination therapy, which resulted in the transition of many ivacaftor monotherapy patients to triple combination therapy. Because of the reduction in recruitment, the study is underpowered to detect associations with clinical outcomes, as prior sample size calculation showed that 30 participants would be needed to evaluate differences in all clinical outcomes with a power of 80% or higher. However, it provides a sample size typically used for understanding variability in pharmacokinetics during typical clinical use.

Conclusion

This study provides evidence that patients taking ivacaftor monotherapy, overall, have more than adequate ivacaftor concentrations in the tissue, as well as in the plasma, while some may have concentrations higher than necessary to improve CFTR function in vivo. Monitoring of modulator concentrations may help identify those patients who may be able to take a reduced dose of ivacaftor while maintaining clinical response.

Abbreviations:

CF

cystic fibrosis

CFTR

Cystic Fibrosis Transmembrane conductance Regulator

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.