Primary Ciliary Dyskinesia: Longitudinal Study of Lung Disease by Ultrastructure Defect and Genotype

Abstract

Rationale: In primary ciliary dyskinesia, factors leading to disease heterogeneity are poorly understood.

Objectives: To describe early lung disease progression in primary ciliary dyskinesia and identify associations between ultrastructural defects and genotypes with clinical phenotype.

Methods: This was a prospective, longitudinal (5 yr), multicenter, observational study. Inclusion criteria were less than 19 years at enrollment and greater than or equal to two annual study visits. Linear mixed effects models including random slope and random intercept were used to evaluate longitudinal associations between the ciliary defect group (or genotype group) and clinical features (percent predicted FEV₁ and weight and height z-scores).

Measurements and main results: A total of 137 participants completed 732 visits. The group with absent inner dynein arm, central apparatus defects, and microtubular disorganization (IDA/CA/MTD) (n = 41) were significantly younger at diagnosis and in mixed effects models had significantly lower percent predicted FEV₁ and weight and height z-scores than the isolated outer dynein arm defect (n = 55) group. Participants with CCDC39 or CCDC40 mutations (n = 34) had lower percent predicted FEV₁ and weight and height z-scores than those with DNAH5 mutations (n = 36). For the entire cohort, percent predicted FEV₁ decline was heterogeneous with a mean (SE) decline of 0.57 (0.25) percent predicted/yr. Rate of decline was different from zero only in the IDA/MTD/CA group (mean [SE], -1.11 [0.48] percent predicted/yr; P = 0.02).

Conclusions: Participants with IDA/MTD/CA defects, which included individuals with CCDC39 or CCDC40 mutations, had worse lung function and growth indices compared with those with outer dynein arm defects and DNAH5 mutations, respectively. The only group with a significant lung function decline over time were participants with IDA/MTD/CA defects.

Keywords: Kartagener syndrome; cilia; respiratory function tests.

Figure 1.

Flowchart outlining the enrolled participants. PCD = primary ciliary dyskinesia; EM = electron micrographs.

Figure 2.

Serial cross-sectional plot of the prevalence of bacteria isolated from respiratory cultures by age category.

Figure 3.

(A) Forest plots of association of percent predicted FEV₁ and weight and height z-scores with ciliary ultrastructural defects: ODA (reference group, n = 55), ODA/IDA (n = 20), IDA/CA/MTD (n = 40), and normal EM (n = 12). Linear mixed effects models including data from all study visits were used to estimate the overall association between clinical feature and defect group. (B) Forest plots of association of percent predicted FEV₁ and weight and height z-scores with primary ciliary dyskinesia mutations. DNAH5 (reference group, n = 36), CCDC39 and CCDC40 (n = 33), DNAH11 (n = 11), other ODA (n = 15), and ODA/IDA (n = 17). Linear mixed effects models including data from all study visits were used to estimate the overall association between clinical feature and genetic defect group. CA = central apparatus; EM = electron micrographs; IDA = inner dynein arm; MTD = microtubular disorganization; ODA = outer dynein arm.

Figure 4.

Estimated mean annual change in percent predicted FEV₁ (ppFEV₁) for each ciliary ultrastructural defect. (A–D) Change in lung function with age for each individual in each defect group, with the estimated slope for that defect group superimposed. (A) ODA, ppFEV₁ = 95.54 − 0.73 × age (yr). (B) ODA + IDA, ppFEV₁ = 88.22 −  0.48 × age (yr). (C) IDA/CA/MTD, ppFEV₁ = 85.54 − 1.1 × age (yr). (D) Normal EM, ppFEV₁ = 85.77 + 0.29 × age (yr). Estimated slopes are from linear mixed effects models including ultrastructure defect group, age in years, and the interaction between them as fixed effects. Individuals with the bold lines are those who were diagnosed based on an ultrastructural defect without a corresponding genetic mutation. For definition of abbreviations, see Figure 3.