PREFUL MRI for Monitoring Perfusion and Ventilation Changes after Elexacaftor-Tezacaftor-Ivacaftor Therapy for Cystic Fibrosis: A Feasibility Study

Abstract

Purpose

To assess the feasibility of monitoring the effects of elexacaftor-tezacaftor-ivacaftor (ETI) therapy on lung ventilation and perfusion in people with cystic fibrosis (CF), using phase-resolved functional lung (PREFUL) MRI.

Materials and Methods

This secondary analysis of a multicenter prospective study was carried out between August 2020 and March 2021 and included participants 12 years or older with CF who underwent PREFUL MRI, spirometry, sweat chloride test, and lung clearance index assessment before and 8–16 weeks after ETI therapy. For PREFUL-derived ventilation and perfusion parameter extraction, two-dimensional coronal dynamic gradient-echo MR images were evaluated with an automated quantitative pipeline. T1- and T2-weighted MR images and PREFUL perfusion maps were visually assessed for semiquantitative Eichinger scores. Wilcoxon signed rank test compared clinical parameters and PREFUL values before and after ETI therapy. Correlation of parameters was calculated as Spearman ρ correlation coefficient.

Results

Twenty-three participants (median age, 18 years [IQR: 14–24.5 years]; 13 female) were included. Quantitative PREFUL parameters, Eichinger score, and clinical parameters (lung clearance index = 21) showed significant improvement after ETI therapy. Ventilation defect percentage of regional ventilation decreased from 18% (IQR: 14%–25%) to 9% (IQR: 6%–17%) (P = .003) and perfusion defect percentage from 26% (IQR: 18%–36%) to 19% (IQR: 13%–24%) (P = .002). Areas of matching normal (healthy) ventilation and perfusion increased from 52% (IQR: 47%–68%) to 73% (IQR: 61%–83%). Visually assessed perfusion scores did not correlate with PREFUL perfusion (P = .11) nor with ventilation-perfusion match values (P = .38).

Conclusion

The study demonstrates the feasibility of PREFUL MRI for semiautomated quantitative assessment of perfusion and ventilation changes in response to ETI therapy in people with CF.

Keywords: Pediatrics, MR–Functional Imaging, Pulmonary, Lung, Comparative Studies, Cystic Fibrosis, Elexacaftor-Tezacaftor-Ivacaftor Therapy, Fourier Decomposition, PREFUL, Free‐Breathing Proton MRI, Pulmonary MRI, Perfusion, Functional MRI, CFTR, Modulator Therapy, Kaftrio

Clinical trial registration no. NCT04732910

Supplemental material is available for this article.

© RSNA, 2024

Summary

This feasibility study demonstrates phase-resolved functional lung, or PREFUL, MRI as an adequate tool to monitor ventilation and perfusion improvement in response to therapy in individuals with cystic fibrosis.

Key Points

• ■ Phase-resolved functional lung (PREFUL) MRI demonstrated improved regional ventilation and perfusion following elexacaftor-tezacaftor-ivacaftor (ETI) therapy in individuals with cystic fibrosis: Ventilation defect percentage of regional ventilation decreased from 18% (IQR: 14%–25%) to 9% (IQR: 6%–17%) (P = .003) and perfusion defect percentage from 26% (IQR: 18%–36%) to 19% (IQR: 13%–24%) (P = .002).

• ■ Areas of matching normal (healthy) ventilation and normal perfusion measured with PREFUL MRI increased from 52% (IQR: 47%–68%) to 73% (IQR: 61%–83%) (P < .001) 2–4 months after ETI treatment.

• ■ Changes in dynamic ventilation heterogeneity measured with PREFUL MRI correlated with lung clearance index changes (P = .03) after ETI treatment.

Introduction

Cystic fibrosis (CF) is a progressive multiorgan disease and the most common lethal autosomal recessive pulmonary hereditary disorder (1,2). It is caused by various mutations of the CF transmembrane conductance regulator (CFTR) gene, which encodes a chloride-conducting transmembrane channel (3). Dysfunction of this channel causes defective transepithelial chloride secretion for airway surface hydration, resulting in manifold changes of lung structure, ventilation, and perfusion (4–6). Triple combination therapy of the two CFTR correctors, elexacaftor and tezacaftor, and the potentiator, ivacaftor (all, Vertex Pharmaceuticals), henceforth abbreviated ETI, addresses the functional defects associated with the most common mutation, p.Phe508del, also known as F508del. ETI has been shown to have unprecedented clinical benefits in people with CF aged 12 years and older with at least one copy of the F508del allele (7–11). These include improvements in forced expiratory volume in 1 second percent predicted (ppFEV₁), body mass index, and self-reported respiratory symptoms (9,10).

ETI also showed positive effects on ventilation, assessed by multiple-breath washout measurements, and on morphologic changes such as bronchial wall thickening and mucus plugging assessed with MRI in participants 12 years and older (12). Interestingly, in the parent study of the current analysis, lung perfusion, which is expected to improve with improvements in ventilation according to the Euler-Liljestrand reflex, did not show improvement (12). In this multicenter parent study, treatment response was measured using the Eichinger score, which is based on a semiquantitative visual assessment of MRI (13), and regional ventilation (RVent) was not evaluated. A recent retrospective study in 24 children using matrix pencil MRI demonstrated improvement in ventilation and perfusion as response to ETI (14). Another study using dynamic perfluorinated gas MRI showed improved ventilation parameters 28 days after initiation of treatment with ETI in people with CF (15).

MRI techniques with inhaled hyperpolarized noble gases, such as helium 3 and xenon 129, provide regionally specific information on lung ventilation with the drawback of complicated accessibility, high costs, and breath-hold maneuvers necessary for acquisition (15–18). Phase-resolved functional lung (PREFUL) MRI is a validated proton-based free-breathing MRI technique that does not depend on gas inhalation or intravenous gadolinium-based contrast agents (19–21). It is based on the Fourier decomposition approach (22) but is technically different as it encompasses less strict low-pass and high-pass filtering and additional image sorting to enhance ventilation and perfusion dynamics analysis (23). A similar sorting approach is also used in self-gated non–contrast-enhanced functional lung (or, SENCEFUL) imaging; however, this method requires sequences with non–phase-encoded direct current signal-gating capabilities, as sorting is performed for individual k-space lines (24). In contrast, PREFUL MRI can be performed with conventional spoiled gradient-echo sequences, contributing to the wide availability of this method, as demonstrated in a multicenter setting (25,26), and across different sites and MR field strengths (27–29). In addition to patient-friendly acquisition, PREFUL MRI is particularly suited for pediatric examinations (18,30). Thus far, PREFUL MRI has been evaluated as an end point in a clinical trial where it was able to help detect improved ventilation and perfusion in response to inhaler therapy in patients with chronic obstructive pulmonary disease (31).

Graeber et al (12) previously reported improvements in lung clearance index at 2.5% of its starting concentration (LCI_2.5) and MRI morphology scores in people with CF after therapy with ETI in a larger multicenter cohort. The present study assesses RVent and perfusion in response to ETI in a subset of participants by using semiautomated PREFUL MRI, which was not evaluated in the parent study.

Specifically, we aimed to investigate if PREFUL MRI–derived quantitative ventilation and perfusion measures can assess changes in people with CF after initiation of ETI therapy and how these measures relate to established clinical and MRI-derived parameters.

Materials and Methods

Study Design and Participants

This study is a secondary analysis of a single-center subsample (n = 23) of a previously published multicenter study carried out between August 2020 and March 2021 (n = 91) (12). The prospective, observational single-center study was conducted with the approval of the ethics committees of the Hannover Medical School (approval no. 8922_BO_S_2020). Written informed consent was obtained from all participants included in the study and from their parents or legal guardians, as appropriate. Participants included were at least 12 years old, compound heterozygous for F508del and a minimal function mutation or homozygous for F508del, and naive to treatment with ETI. Participants needed to consent to adhere to a stable medication regimen including ETI according to patient labeling and prescription information for the duration of study participation. Exclusion criteria were prior treatment with ETI, acute respiratory infection or pulmonary exacerbation at baseline or follow-up, history of solid organ transplantation, pregnancy, claustrophobia, and metallic implants incompatible with MRI.

The following methods were applied: ppFEV₁ and midexpiratory flow at 25% of forced vital capacity (MEF25) as outcome measures of spirometry, sweat chloride test, anthropometry, LCI_2.5 by nitrogen washout (32,33), and PREFUL MRI assessed at baseline and 8 to 16 weeks after initiation of therapy. Due to personal reasons, one participant underwent posttreatment evaluation only 28 weeks after initiation of therapy.

Participants were treated with the approved and recommended dose of 200 mg elexacaftor and 100 mg tezacaftor every 24 hours in combination with 150 mg ivacaftor every 12 hours as described previously (12).

PREFUL MRI

Non–contrast-enhanced two-dimensional PREFUL MR images were acquired during free breathing with a 1.5-T MRI scanner (Magnetom Avanto; Siemens Healthineers) at baseline and 8 to 16 weeks after initiation of therapy. Depending on thorax size, between five and 13 coronal sections, covering the entire lung, were acquired per participant. A spoiled gradient‐echo sequence with the following settings was used: field of view, 450 × 450 mm²; matrix size, 112 × 112 (interpolated to 224 × 224); section thickness, 15 mm; no gap between sections; echo time, 0.81 msec; repetition time, 3 msec; flip angle, 5°; bandwidth, 1490 Hz/pixel; generalized autocalibrating partially parallel acquisitions (or, GRAPPA) with acceleration factor R = 2; and acquisition of 24 autocalibration lines. A total of 250 images per section with a temporal resolution of 160 msec were obtained.

For PREFUL postprocessing, all images were registered toward one fixed image in the intermediate lung position using a group-oriented registration approach (34). Registration was performed with the freely available Advanced Normalization Toolkit (35). Lung parenchyma was segmented with a fully automated U-Net convolutional neuronal network (CNN) segmentation. Sections with suspected strong partial volume artifacts were automatically excluded by a CNN (36). Manual resegmentation was performed by a radiologist with 12 years of experience in respiratory radiology (M.D.), if necessary, to ensure all lung parenchyma was included (36). All PREFUL images were resegmented manually once for quality control before diagnostic reading without any adjustment between visual and semiautomated reading. Most common resegmentation adjustments were made to ensure differentiation between vessels and lung parenchyma.

An automated algorithm placed a region of interest in the pulmonary artery, the aorta, and the heart. PREFUL postprocessing was performed as reported previously, and thus the following parameters were obtained: RVent, flow-volume loop correlation metric (FVL_CM), and quantified perfusion (QQuant) (21,23,37,38). FVL_CM is a two-dimensional FVL correlation map based on a voxelwise correlation of FVL values with an individualized reference FVL as measure of ventilation dynamics (39). RVent voxels having a value lower than the 90th percentile multiplied by 0.3 and FVL_CM values below 90% were classified as ventilation defect.

Ventilation defect percentage (VDP) of RVent (VDP_RVent) images were acquired in end-inspiratory and end-expiratory positions and therefore only contain static RVent information of the breathing cycle. VDP_FVL-CM was based on a two-dimensional FVL correlation map and created a defect percentage map of dynamic ventilation over an entire breathing cycle. VDP_combined combined dynamic FVL_CM with static RVent VDP_RVent values and created a combined dynamic and static defect ventilation percentage map of the lung.

Perfusion was quantified by estimating the voxelwise proton density and calculating the median signal decay toward the steady state as a reference for the perfusion amplitude (28). Areas with QQuant values under the threshold (90th percentile ∙ 0.3, which was chosen empirically) were considered as perfusion defect (40).

Perfusion defect and VDPs were then calculated in relation to the lung volume covered in the analyzed sections.

A detailed description of PREFUL parameter abbreviations is given in Table S1.

Eichinger MRI Morphology and Perfusion Score

The following morphologic MRI sequences were performed: T2-weighted half-Fourier acquisition single-shot turbo spin-echo sequences in axial and coronal orientation, acquired during breath-hold maneuvers; T2-weighted turbo spin-echo sequences in periodically rotated overlapping parallel lines with enhanced reconstruction (or, PROPELLER) technique (BLADE; Siemens Healthineers), acquired in axial and coronal orientation by using respiratory navigator triggering; and T1-weighted volumetric interpolated breath-hold examination sequences, acquired in axial and coronal orientation.

Morphologic MRI sequences and functional PREFUL MRI perfusion maps of the entire lung were visually assessed for abnormalities in lung morphology and perfusion. A dedicated morphofunctional MRI score was used, composed of five morphologic subscores and one perfusion subscore, as follows: (a) bronchial wall thickening and/or bronchiectasis, (b) mucus plugging, (c) abscesses and/or sacculations, (d) consolidations, (e) special findings, such as pleural effusion, and (f) changes in lung perfusion as previously described (12,13). The Eichinger score evaluates perfusion at dynamic contrast-enhanced MRI and was adapted to the non–contrast-enhanced methods applied for this study. MR images were read independently by three different readers (M.D., D.M.R., J.V.C.), each with more than 10 years of experience in lung MRI and blinded to participant clinical characteristics and study visit as previously described and added in Appendix S1 (12). For visual evaluation, morphologic images and perfusion maps were simultaneously available per participant.

Multiple-Breath Washout

The Exhalyzer D system (Eco Medics) was used for multiple-breath washout testing to determine LCI_2.5. Resident nitrogen was washed out with 100% oxygen from the lungs using a mouthpiece as interface (32). All measurements were evaluated centrally using Spiroware 3.3.1 (Eco Medics) (33,41,42). The upper limit of normal was determined as 7.1 (42).

Statistical Analysis

The Kolmogorov-Smirnov test revealed that the data were not normally distributed; thus, data are presented as medians and first quartiles of the 25th percentile and third quartiles of the 75th percentile. Clinical parameters and PREFUL values before and after therapy with ETI were compared with a Wilcoxon signed rank test. Using prior repeatability data in patients with chronic obstructive pulmonary disease, a power of 0.8, and an α level of .05, a minimum detectable difference of 6.1% VDP_FVL-CM was calculated. The correlation of change values of clinical parameters and Eichinger score with PREFUL data were calculated as the Spearman ρ correlation coefficient (19). The estimate of the Spearman correlation 95% CI was carried out according to Fieller, Hartley, and Pearson. Statistical analysis was carried out by one author (M.D.) using IBM SPSS Statistics version 27. A P value less than .05 indicated statistical significance.

Results

Characteristics of the Study Sample

A total of 26 participants were eligible for study inclusion. Two participants had to be excluded from the study as MRI could not be carried out because of claustrophobia. Another participant had to be excluded as follow-up MRI was declined for unknown reasons (Fig 1), resulting in a total of 23 participants (median age, 18 years [IQR: 14–24.5 years]; 13 female, 10 male) with a complete analysis set of MRI examinations at baseline and after follow-up 8–16 weeks (median: 16 weeks [IQR: 14–16.5]) after initiation of ETI, with one participant exceeding the interval by 5 weeks but still included.

Figure 1:

Flowchart of participants included in the study. A total of 23 participants with a full analysis set of phase-resolved functional lung (PREFUL) MRI at baseline and follow-up, 21 with lung clearance index (LCI) of 2.5% before and after initiation of treatment with elexacaftor-tezacaftor-ivacaftor, were included in the study. MBW = multiple-breath washout.

Demographics and clinical characteristics are presented in Table 1. All 23 participants had preserved lung function (median ppFEV₁: 89% [IQR: 75%–100%]). LCI_2.5 was determined in 21 participants because of compliance difficulties during breath washout maneuvers in two participants (Fig 1). Parameters for spirometry, sweat chloride test, anthropometry, LCI_2.5, and MRI global and morphology scores improved significantly under therapy with ETI in the selected subset of 23 participants (Tables 2, 3; Fig 2) (12). Perfusion defects as part of Eichinger MRI score showed a significant improvement of −1 point (IQR: −3 to 0) when analyzed visually for this subset of participants (P = .03) (Table 3).

Table 1:

Clinical Characteristics of Study Participants

Table 2:

Change in Clinical Parameters and Lung Function Tests after Initiation of Therapy with ETI

Table 3:

Change in Eichinger Score after Initiation of Treatment with ETI

Figure 2:

Images from T2-weighted turbo spin-echo sequences in periodically rotated overlapping parallel lines with enhanced reconstruction (or, PROPELLER) technique (BLADE; Siemens Healthineers) in coronal orientation in an 18-year-old male participant with cystic fibrosis. In the right upper lobe, bronchial wall thickening and mucus plugging are visible before initiation of elexacaftor-tezacaftor-ivacaftor (ETI, left), with well-visible reduction of morphologic findings after initiation of therapy with ETI (right). These morphologic images correspond to ventilation and perfusion defects in the right upper lobe in the phase-resolved functional lung (or, PREFUL) maps (Fig 3).

Change in PREFUL Parameters after Initiation of Treatment with ETI

PREFUL ventilation and perfusion parameters showed significant improvement after treatment with ETI in measurements including the entire lung (Table 4; Figs 3, 4). Ventilation defects VDP_RVent decreased from 18% (IQR: 14%–25%) to 9% (IQR: 6%–17%), reflecting a decrease of 50% (P = .003). Perfusion defects decreased from 26% (IQR: 18%–36%) to 19% (IQR: 13%–24%), which corresponds to a 27% reduction (P = .002) (Table 4, Fig 4). A full cardiac and a full ventilation cycle before and after ETI in the same patient as in Figure 2 are shown in Movies 1–4. Median values of perfusion (QQuant) improved by 13.5% from 47 mL/min/100 mL (IQR: 41–59) to 54 mL/min/100 mL (IQR: 46–72) (P = .003), and dynamic ventilation heterogeneity (FVL_CM) improved by 6.6% from 91% (IQR: 86%–95%) to 97% (IQR: 92%–98%) (P < .001) (Table 4) (39). Areas of matching normal (healthy) ventilation and perfusion increased by 40.4% from 52% (IQR: 47%–68%) to 73% (IQR: 61%–83%) (P < .001), whereas areas of matching ventilation and perfusion defect decreased from 9% (IQR: 5%–20%) to 2% (IQR: 1%–12%) (P < .001) (Table 4, Fig 3).

Table 4:

Change in Phase-resolved Functional Lung MRI Parameters of Entire Lung before and 8–16 Weeks after Initiation of Therapy with ETI

Figure 3:

Phase-resolved functional lung (PREFUL) MRI ventilation and perfusion maps in an 18-year-old male participant with cystic fibrosis at baseline (quantified perfusion [QQuant]: 47 mL/min/100 mL, flow-volume loop correlation metrics [FVL_CM]: 92%, ventilation defect percentage of FVL correlation metrics [VDP_FVL-CM]: 14%, ventilation perfusion match [VQM] healthy: 70%, VQM defect: 9%) and 13 weeks after initiation of treatment with elexacaftor-tezacaftor-ivacaftor (ETI) (QQuant: 57 mL/min/100 mL, FVL_CM: 98%, VDP_FVL-CM: 4%, VQM healthy: 79%, VQM defect: 1%). Contrast-free PREFUL maps allow for visual detection of ventilation and perfusion defects and their resolution under treatment with ETI. Note improved perfusion (upper row) and ventilation (middle row) of right upper lobe resulting in decrease of ventilation-perfusion mismatch (lower row), seen as ventilation defect (blue) with normal perfusion before therapy and normalized ventilation with normal perfusion (dark green) after therapy with ETI. Also, matching areas with combined reduced ventilation and perfusion (purple before therapy) resolved after therapy with ETI. QDP = perfusion defect percentage.

Figure 4:

Change in phase-resolved functional lung (PREFUL) parameters at baseline and after therapy with elexacaftor-tezacaftor-ivacaftor (ETI) in all participants. Effects of ETI on PREFUL ventilation parameters of (A) ventilation defect percentage of regional ventilation (VDP_RVent), (B) flow-volume loop correlation metrics (FVL_CM), (C) VDP of FVL_CM (VDP_FVLCM), and (D) VDP combining RVent and FVL (VDP_combined) and on PREFUL perfusion parameters (E) quantified perfusion (QQuant) and (F) quantified defect percentage of quantified perfusion (QDP_QQuant) in participants with cystic fibrosis at baseline and 8–16 weeks after initiation of ETI therapy. P < .05 was considered statistically significant.

Movie 1:

Download video file ^{(2.1MB, mp4)}

A full cardiac cycle in a midthoracic position before treatment with ETI in an 18-year-old male participant with cystic fibrosis is shown. A perfusion defect can be noted in the upper right lobe (dark red/black).

Movie 2:

Download video file ^{(2.4MB, mp4)}

A full ventilation cycle in a midthoracic position before treatment with ETI in an 18-year-old male participant with cystic fibrosis is shown. A ventilation defect can be noted in the upper right lobe (dark gray/black).

Movie 3:

Download video file ^{(2.4MB, mp4)}

A full cardiac cycle in a midthoracic position after treatment with ETI in the same 18-year-old male participant with cystic fibrosis as in Movie 1 is shown. The former perfusion defect in the right upper lobe has fully disappeared.

Movie 4:

Download video file ^{(2.8MB, mp4)}

A full ventilation cycle in a midthoracic position after treatment with ETI in the same 18-year-old male participant with cystic fibrosis as in Movie 2 is shown. The former large ventilation defect in the right upper lobe has decreased in size and only a small peripheral area of impaired ventilation in the upper lobe is now visible (dark gray).

Correlation of Change Values of PREFUL MRI, Eichinger Score, and Clinical Parameters

Mild to moderate correlations were detected between change values of PREFUL MRI ventilation parameters and changes of lung function parameters (Table S2): Change in FVL_CM showed moderate correlation with change in LCI_2.5 (ρ = −0.45 [95% CI: −0.73, −0.03]) (Fig 5), change in VDP_FVL-CM correlated with change in ppFEV₁ (ρ = −0.51 [95% CI: −0.77, −0.11) (Fig 6), and change in FVL_CM correlated with change in MEF25 (ρ = 0.44 [95% CI: 0.02, 0.73) (Fig 7).

Figure 5:

Change in flow-volume loop correlation metrics (FVL-CM) shows moderate correlation with change in lung clearance index (LCI) (ρ = −0.45 [95% CI: −0.73, −0.03]).

Figure 6:

Change in ventilation defect percentage of flow-volume loop correlation metrics (VDP-FVL-CM) shows moderate correlation with change in forced expiratory volume in 1 second percent predicted (ppFEV1) (ρ = −0.51 [95% CI: −0.77, −0.11]).

Figure 7:

Change in flow-volume loop correlation metrics (FVL-CM) shows moderate correlation with change in midexpiratory flow at 25% of forced vital capacity (MEF25) (ρ = 0.44 [95% CI: 0.02, 0.73]).

There was no correlation of change values of the PREFUL MRI perfusion parameters with changes of lung function parameters nor with changes of Eichinger global or subscores, including perfusion score (all P > .21) (Table S3). When comparing change values of Eichinger score with change values of lung function parameters, only MEF25 correlated with the subscore of bronchial wall thickening (ρ = 0.47).

Discussion

This study demonstrates that semiautomated quantitative PREFUL MRI can serve as a reliable tool to detect improvements in RVent and perfusion in people with CF in response to CFTR modulation therapy with ETI. Corresponding to this finding, changes in several PREFUL ventilation and perfusion parameters correlated with changes in the lung function parameters LCI_2.5, ppFEV₁, and MEF25 and with the Eichinger perfusion score.

Semiautomated quantitative PREFUL in comparison to visual semiquantitative PREFUL has the advantage of high repeatability and lower interreader variability as lung parenchyma is “read” with a fully automated U-Net CNN segmentation (19,31,36). Our data show that PREFUL MRI permits evaluation of ventilation and perfusion in the same examination, including anatomic resolution to detect and follow up areas with normal or reduced ventilation and/or perfusion. The possibility to calculate matching ventilation and perfusion areas as well as regions of ventilation-perfusion mismatch may serve as a useful end point for monitoring response to therapy in people with CF.

Ventilation assessment is not part of the Eichinger scoring system, as this score evaluates perfusion at dynamic contrast-enhanced MRI. We adapted this score to the non–contrast-enhanced method of PREFUL, which also assesses ventilation. Commonly, ventilation inhomogeneity is assessed by LCI_2.5 and indirectly by spirometry. These were both performed in all participants and showed significant improvement (P < .001) following ETI treatment. Similarly to LCI_2.5, ppFEV₁, and MEF25 measurements, PREFUL ventilation parameters improved significantly after initiation of therapy with ETI (all P ≤ .003). Dynamic ventilation heterogeneity (FVL_CM) improved from 91% to 97%, approaching nearly normal levels after treatment with ETI, a milestone of considerable importance for people with CF. Corroborating the capacity of PREFUL parameters to accurately assess ventilation and perfusion improvement after initiation of ETI, there were moderate correlations of the ventilation and perfusion PREFUL parameters with LCI_2.5 and spirometry measurements. As an important asset of PREFUL as an imaging tool, our data show PREFUL maps permit anatomic resolution of changes in ventilation, which is not possible with LCI_2.5 or spirometry.

In contrast to a previous analysis of a larger multicentric cohort (12), the present study shows significant improvement not only in ventilation but also in lung perfusion in people with CF after initiation of ETI therapy assessed with quantitative PREFUL MRI. A recently published retrospective study in 24 children has observed improvements in ventilation and perfusion as a response to ETI using matrix pencil MRI (14). Graeber et al (12) have previously described significant improvements in lung ventilation, assessed by LCI_2.5, and improvements of thoracic MRI morphology, assessed by semiquantitative visual scoring, in people with CF after initiation of ETI. However, the authors found no improvement of pulmonary perfusion in this large multicentric cohort (91 participants) (12), nor did another study group reporting on a small number of people with CF after treatment with ETI (43). In contrast to the results of the larger parent study, improvement of perfusion could be detected by semiquantitative visual assessment scoring according to Eichinger in our subgroup. These discrepancies might result from different aspects that are not mutually exclusive but might be additive. First, for perfusion assessment, one-fourth of the participants of the parent study underwent contrast-free PREFUL imaging, which constitutes our subgroup, and three-fourths underwent contrast-based imaging of perfusion (12). Second, semiautomated quantitative evaluation of the same PREFUL MR images of the subgroup confirmed a significant improvement of regional perfusion (QQuant) by 13.5% from 47 to 54 mL/min/100 mL 8–16 weeks after initiation of therapy. This additional semiautomated pixel-by-pixel control of visual evaluation could not be carried out in the parent study and may have revealed improvements too subtle to detect visually. Another reason for this difference might be preserved lung function of our subgroup with ppFEV₁ of 89% (IQR: 75%–100%) as compared with the parent study cohort, with ppFEV₁ of 74.3% (IQR: 54.8%–90.4%) (F508del/minimal function) and 81% (IQR: 63.7%–95.2%) (F508del homo). With more preserved lung function, perfusion defects might improve more quickly in response to therapy with ETI.

Furthermore, by resorting to PREFUL MRI, we were able to demonstrate improvements in regional ventilation-perfusion matching after ETI therapy. These responses were expected, as a decrease in inflammation and bronchial mucus plugging because of ETI were assumed to lead to improved ventilation, which in turn improves regional parenchymal perfusion, likely because of the Euler-Liljestrand reflex (44). However, the correlation of these mechanisms has not yet been formally assessed. Our data indicate that regional perfusion improvement seems to be less marked compared with the RVent changes 2–4 months after ETI therapy. Ventilation impairment in people with CF is most likely more dependent on mechanical obstruction because of mucus plugging and lumen narrowing through bronchial wall thickening, which Graeber et al (12) previously demonstrated to rapidly improve in response to ETI. These findings might also indicate perfusion reacts slower than ventilation to ETI therapy. In that line, perfusion might be more affected by residual parenchymal inflammation or postinflammatory changes 2–4 months after ETI therapy, compared with the more immediate effects of ETI on ventilation. Persistent remodeling of the pulmonary microvasculature because of chronic inflammation might be a contributing factor (43,44). So far, studies do not yet permit definitive conclusions regarding the effect of ETI on chronic inflammation, neither systemically nor by imaging alone, as persistent inflammation might continue to drive remodeling of the pulmonary microvasculature. Thus, further studies including MRI and other biomarkers of chronic inflammation with longer follow-up intervals after initiation of therapy are needed to better understand the relationship between perfusion and chronic inflammation.

Quantitative PREFUL MRI could help detect ventilation and perfusion improvement in people with CF after initiation of therapy with ETI. The presented data substantially expand the knowledge on emerging semiautomated PREFUL MRI–derived quantitative ventilation and perfusion parameters measuring treatment response to ETI in people with CF. Semiautomated PREFUL provides quantitative spatial and anatomic resolution of the lung, enabling therapy surveillance on a functional level, which likely adds value to clinical monitoring of people with CF. Therapy effects could be locally monitored and directed toward most affected areas, that is, physiotherapy to improve atelectasis. Particularly more subtle therapeutic effects or the lack thereof in less advanced CF might be difficult to detect and assess clinically. PREFUL MRI could be used for guidance not only for directed physiotherapy but also in regard to the reduction of supportive therapies such as secretolytics. In addition, PREFUL MRI is contrast agent and radiation free, making it an ideal tool for repetitive examinations in chronically ill patients, and, as carried out in free breathing, especially in children. However, further studies are needed to depict the best-performing parameters of semiautomated quantitative PREFUL MRI.

Our observational design constituted a clear limitation of our study. With regard to the MRI measurements, including PREFUL, we tried to minimize these effects by blinded and semiautomated review. Further limitations of our study were the single-center study design and the relatively small cohort size of only 23 participants. Interestingly, even with this small sample size, we were able to show very good improvements in the majority of PREFUL parameters analyzed, underscoring the sensitivity and thus the potential utility of PREFUL MRI as an end point for clinical response to therapeutic changes in people with CF. This is particularly interesting as the subcohort analyzed presented with well-preserved lung function, indicating less advanced lung disease, where PREFUL MRI was nonetheless able to help identify substantial improvements. As in this study outcome measures were considered exploratory, descriptive P values are presented and no correction for multiple comparison was carried out. Due to the small sample size, we decided against further multiple regression analysis. Another temporary limitation was that PREFUL is currently commercially available for research use only.

In conclusion, PREFUL is a promising tool to monitor dynamic RVent and perfusion, as well as ventilation-perfusion matching, in people with CF undergoing ETI therapy. Further larger multicenter trials are needed to confirm our single-center findings.