Primary ciliary dyskinesia (PCD) is a rare, ciliopathic, autosomal recessive genetic disorder that causes defects in the action of cilia and can be accompanied by situs inversus. The main impact of impaired ciliary function on the respiratory system is the reduced mucus clearance from the lungs, which leads to highly increased susceptibility to chronic recurrent respiratory infections and impairment of pulmonary function¹. To this end, the assessment of the lung ventilation and perfusion in PCD patients and periodic monitoring of the disease progression, is of high importance. In the recent years a novel, non-invasive and contrast-agent free technique called Fourier decomposition (FD) MRI² was proposed for evaluation of regional ventilation and perfusion. The purpose of this work was to assess the pulmonary function in PCD patients using FD MRI in combination with the multiple breath washout (MBW) technique³ and spirometry.

MR measurements

Nine PCD patients (mean age: 11.1 years , 5.9-17.2 years range) were examined. Scans were performed on a 1.5 T MR-scanner (Avanto, Siemens Healthcare, Germany) using a combination of 12-channel body and 24-channel spine matrix coils. Time-resolved sets of coronal images were acquired using ultra-fast steady-state free precession (ufSSFP) sequence⁴ in free-breathing to cover the whole lung volume. The imaging parameters were: TE/TR/TA=0.69/1.57/119 ms, FA=60°, field-of-view=400x400 mm², matrix=128x128 (interpolated to 256x256), slice thickness=12 mm, bandwidth=2056 Hz/pixel, GRAPPA factor=2, imaging rate=4 images/s, 200 images.

MBW and spirometry

All subjects underwent nitrogen-multiple breath washout examination with a commercially available device (ExhalyzerD, EcoMedics, Switzerland). Main outcome parameters representing global ventilation inhomogeneity (VI) was the lung clearance index $$$LCI_{2.5\%}$$$, second moment ratio $$$M_2/M_0$$$; conductive airways VI - $$$S_{cond}$$$; acinary VI - $$$S_{acin}$$$. Subsequently, spirometry was performed using a Jaeger MasterScreen bodyplethysmograph (CareFusion, Hoechberg, Germany). Outcome parameters were $$$FEV_1$$$ (forced expiratory volume in 1 second) and $$$FEF_{25-75}$$$ (forced expiratory flow 25–75 %) as a marker for obstruction in the central airways.

Image postprocessing

Nonrigid image registration was used to correct for the respiratory motion⁵. Subsequently, the first image in every registered dataset was semi-automatically segmented and projected along the time axis. Fourier decomposition was applied to detect and spectrally separate periodic changes of the lung proton density caused by the respiratory and cardiac cycles. Maps of fractional ventilation $$$FV$$$⁶, perfusion $$$Q$$$⁷ and blood arrival time $$$BAT$$$⁸ were generated.

Automated evaluation of perfusion and ventilation defects

Distributions of voxel values on the segmented $$$FV$$$ and $$$Q$$$ images were calculated for every slice position. Voxel values lower than thresholds defined as $$$0.75{\cdot}{median}\{FV,Q\}$$$ of the voxel distribution were regarded as regions with impaired ventilation or perfusion. The ratio of the lung volume with impaired perfusion $$$R_Q$$$ and fractional ventilation $$$R_{FV}$$$ and the whole lung volume was calculated for every patient.

Statistical analysis

Parameters measured using MBW and spirometry were correlated with percentage of impaired perfusion and fractional ventilation obtained using FD MRI in all patients using Pearson's correlation coefficient. P-values<0.05 were considered statistically significant.

Figure 1 shows examplar native coronal ufSSFP, fractional ventilation, perfusion and blood arrival time images obtained in a 11 year old male PCD patient using FD MRI. Regions with reduced fractional ventilation, perfusion and increased blood arrival time located in the basal parts of both lungs are marked with white arrows. Fractional ventilation and perfusion images in 8 year old male PCD patient are presented in figure 2. Regions with reduced fractional ventilation and perfusion (top row) located in the sixth segment of the left lung and basal segments of both lungs were automatically detected and masked on the morphological chest images (bottom row).

The median $$$R_{FV}$$$ in the patient group was 0.17 (max=0.23, min=0.08), the median $$$R_{Q}$$$ was 0.15 (max=0.25, min=0.07). Statistically significant correlations between $$$R_{FV}$$$ and MBW, spirometry parameters were found for $$$S_{cond}$$$ (0.80,P=0.009) and $$$FEF_{25-75}$$$ (0.71,P=0.032). No significant correlations were found between MBW, spirometry and $$$R_{Q}$$$.

In this work we show the feasibility of automatic evaluation of regional lung function using the contrast-media-free FD MRI in patients with PCD. Altough the patient sample in this preliminary study was small, it was possible to find significant correlations between FD MRI and the parameters obtained using multiple breath washout technique and spirometry. Nevertheless, further evaluation in a larger patient group is needed to achieve adequate statistical power.

FD MRI can be used for detection of early and subtle functional impairment. Thank to low compliance requirements of FD MRI, the examinations were successfully performed even in the youngest patients. Hence, the technique can be recommended for non-invasive monitoring of therapy response and longitudinal follow-ups.