Abstract
Purpose
Pulmonary disease is the major cause of morbidity and mortality in osteogenesis imperfecta (OI). We investigated the contribution of intrinsic lung factors to impaired pulmonary function in children and young adults with OI types III, IV, VI.
Methods
Patients with type III (n=8), IV (n=21), VI (n=5), VII (n=2) or XIV (n=1) OI (mean age 23.6 years) prospectively underwent pulmonary function tests (PFTs) and thoracic CT and radiographs.
Results
PFT results were similar using arm span or ulnar length as height surrogates. PFTs were significantly lower in type III than type IV or VI OI. All patients with type III and half of type IV OI had lung restriction; 90% of patients with OI had reduced gas exchange. Patients with COL1A1 variants had significantly lower forced expiratory flow (FEF)25%–75% compared with those with COL1A2 variants. PFTs correlated negatively with Cobb angle or age. CT scans revealed small airways bronchial thickening (100%, 86%, 100%), atelectasis (88%, 43%, 40%), reticulations (50%, 29%, 20%), ground glass opacities (75%, 5%, 0%), pleural thickening (63%, 48%, 20%) or emphysema (13%, 19%, 20%) in type III, IV or VI OI, respectively.
Conclusion
Both lung intrinsic and extrinsic skeletal abnormalities contribute to OI pulmonary dysfunction. Most young adult patients have restrictive disease and abnormal gas exchange; impairment is greater in type III than type IV OI. Decreased FEF25%–75% and thickening of small bronchi walls indicate a critical role for small airways. Lung parenchymal abnormalities (atelectasis, reticulations) and pleural thickening were also detected. Clinical interventions to mitigate these impairments are warranted.
Trial registration number
INTRODUCTION
Osteogenesis imperfecta (OI), a heterogeneous group of inherited skeletal dysplasias, is a collagen-related disorder.1–3 The hallmarks of OI are bone fragility, leading to ease of fracture, skeletal deformities and growth deficiency.1 2 OI is also a generalised connective tissue disorder with defects in extraskeletal organs.1 Notably, respiratory disease is a leading cause of morbidity and mortality in OI, either directly or indirectly, and respiratory symptoms negatively impact quality of life.4–7 An examination of patients with OI in a comprehensive Danish registry revealed cardiopulmonary dysfunction as the highest cause of morbidity and mortality in OI.5
For many years, the prevailing view was that restrictive lung disease in OI was secondary to scoliosis and chest wall deformity, and eventually led to cardiac failure. Pulmonary function tests (PFTs) in adult patients with OI dropped sharply when scoliosis curves exceeded 30°.8 In addition to respiratory failure and chronic ventilatory defects secondary to severe vertebral and rib cage deformities, individuals with OI are susceptible to respiratory infections and bronchiectasis.1 4 5 The first clear indications of bone-independent pulmonary disease in OI came from both genetically modified mice with a dominantly inherited collagen defect, the Aga2 mouse and children with types III or IV OI.9 In a cohort of 46 children with OI, we reported that PFT-derived total lung capacity (TLC), vital capacity and forced vital capacity (FVC) decline with age. Importantly, when children without scoliosis were analysed separately, significant declines in PFTs with age occurred. Later, a murine study of recessive type VII OI demonstrated distortion of normal lung architecture in the Crtap−/− mouse.10 11
Thus, lung function measurements and an understanding of the mechanism of lung dysfunction in OI are important for the clinical management of patients with OI and as a basis for devising approaches to improve care. Given the presence of type I collagen in lung parenchyma and airways,12 it is perhaps not surprising that intrinsic abnormalities would be part of OI lung pathology.
Here, we expand the pulmonary phenotype of children and extend our investigation to young adults with different OI types. Most patients in our cohort had classical OI, with moderately severe type IV (OMIM 166220) or severe progressive type III OI (OMIM 259420), which are characterised by defective structure or quantity of type I collagen.1 13 14 Some participants had type VI (OMIM 613982), type VII (OMIM 610682) or type XIV (OMIM 610682) OI, recessive forms caused by null mutations in SERPINF1, CRTAP or TMEM38B, respectively.1 3 15 16 Types VII and XIV have defects in collagen post-translational modification. We hypothesised that abnormalities of pulmonary function in OI are not entirely due to scoliosis or abnormal thoracic shape, but rather, that there is an intrinsic pulmonary parenchymal component to this clinically significant secondary feature of OI. We identified structural abnormalities in the lungs and pleura independent of bony thoracic defects. This prospective observational cohort study further defines functional and structural abnormalities of the airways, pulmonary tissue or pleura in patients with OI and broadens our understanding of clinical features of lung disease in OI.
MATERIALS AND METHODS
Study population
All subjects were diagnosed with OI as described.17 Genotyping was performed for all subjects (see online supplemental table 1). Thirty-seven individuals with moderate-to-severe OI were evaluated prospectively at the NIH Clinical Center from 2018 through 2022 (Clinical Trials NCT03575221; A Natural History of the Collagen-Related Disorder Osteogenesis Imperfecta and the Genotype-Phenotype Correlation). A respiratory questionnaire (see online supplemental material) was administered by SKT to each subject or parents of children orally or in writing about histories of smoking, prior respiratory infections, chronic respiratory disease diagnoses, cough, use of walking aids or dyspnoea. Since these were patient-reported results, they were necessarily subjective and reflected the voices of the patients about their respiratory quality of life. Key questionnaire results were correlated with the presence or absence of restrictive lung disease in patients with type IV OI, who comprised the largest subgroup. Due to the complexity and variability of questionnaire responses, correlations were not performed with CT scans.
Pulmonary function testing
Spirometry, lung volumes by nitrogen washout and diffusion capacity studies were performed in accordance with American Thoracic Society (ATS)/European Respiratory Society (ERS) standards as described.18 19 PFT results were obtained by an experienced examiner (MB), who determined whether values met standard criteria. Our pulmonary function lab follows ERS/ATS standards (https://gcc02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.thoracic.org%2Fstatements%2Fresources%2Fpft%2FDLCO.pdf&data=05%7C01%7Cmarinij%40cc1.nichd.nih.gov%7C1d9545c8116546cdc26608db40348560%7C14b77578977342d58507251ca2dc2b06%7C0%7C0%7C638174366537393808%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=7jYePwKKEc%2FfzYk%2Fs8BC1mPKMg%2BJ2a24%2F38SzQzx9YU%3D&reserved=0).
We grade diffusion capacity efforts A–D where:
Grade A: Inspitatory volume/vital capacity (VI/VC) ≥90% (or VI/VC >85% and alveolar volume (VA) is within 200 mL and 5% (whichever is greater) of the largest VA from another acceptable effort.
Grade B: VI/VC ≥85%.
Grade C: VI/VC ≥80%.
Grade D: VI/VC ≤80%, breath-holding time (BHT) <8 s or >12 s.
If no grade A manoeuvre is obtained, manoeuvres of grades B–D might still have clinical utility. For these reasons, there may have been tests where an effort <90% of best VC was reported. PFT data determined to be unreliable or unreproducible were not reported. Specifically, five individuals (one with type III OI, two with type IV OI, one with type VI OI and one with type VII OI) were unable to perform some or all PFTs; only reliable values from these individuals were analysed.
Patients with OI typically have significant short stature secondary to bony deformities, such as scoliosis. Short or bowed lower extremities exert less effect on trunk size. Because height is a parameter used to calculate PFT predicted values, OI skeletal deformities can distort PFT calculations that use patient height. To determine whether arm span or ulnar length measurements are comparable surrogates for height for PFT calculations in patients with OI, we compared percentages of predicted PFTs calculated using these height surrogates. Arm span (ie, longest distance measured between tip of third digit in each hand, accounting for upper extremity deformity) and ulnar length (ie, mean distance of three measurements performed with a tape measure between the olecranon process and the midpoint of the styloid process of the same forearm) were used as height surrogates for PFT calculations. Percentages of predicted values for FVC, forced expiratory volume in 1 s (FEV1), forced expiratory flow at 25%–75% of FVC (FEF25%–75%), TLC and adjusted diffusion capacity (DLCOa) were calculated using standardised equations. By convention, normal percentages of predicted values for FVC, FEV1, TLC and DLCOa are >80% of predicted, while for FEF25%–75%, values >65% of predicted are considered to be normal.
Radiographic imaging
Chest CT scans without intravenous contrast were performed as described.20 CT images were reviewed by a pulmonologist with expertise in rare lung diseases who was blinded to each subject’s OI type (BRG). CT images were evaluated for bronchiectasis, atelectasis, reticulations or interstitial lung disease, ground glass opacities, infiltrates, lung nodules, emphysema (ie, low attenuation lung parenchyma), lung cysts, bony chest wall deformities, pleural thickening, pleural effusions or presence of vertebral rods. Thickness of airway walls and outer airway diameters at levels of the trachea, bronchus intermedius and lower lobar subsegmental bronchi were measured using electronic callipers. Lung tissue density was measured in 10 pixels, each with an approximate area of 10 mm2, randomly selected per lobe from regions without visible airways or blood vessels. Hounsfield units in each pixel were used to calculate mean lung tissue density for each lobe and for whole lungs in each individual. Cobb angles were measured from anteroposterior thoracic spine radiographs by two investigators (ANDD and JCM) using the angle tool in the Picture Archiving and Communication System as described (https://raduopaedia.org/articles/cobb-angle?lang=us). Scoliosis was defined as Cobb angle ≥10°.
Statistical analysis
Data are shown as mean±SD. By convention, normal percentages of predicted values for FEV1, FVC, TLC and DLCOa are >80% of predicted, while for FEF25%–75%, values >65% of predicted are considered to be normal.
Statistical analyses were conducted on data from OI types with sufficient n, that is, OI types III versus IV, OI types III versus VI and OI types IV versus VI. Statistical significance was calculated using a two-tailed, unpaired Student’s t-test (GraphPad Prism, Dotmatics, Boston, Massachusetts, USA). Pairwise comparisons for patients with type IV OI with and without restrictive lung disease were performed using Fisher’s exact test. Best-fit lines using simple linear regression analysis to correlate pulmonary function values (FEV1, FVC, FEF25%–75%, TLC, DLCOa percentages of predicated values) with Cobb angle or age were generated (GraphPad Prism). A p value of <0.05 was considered to be statistically significant.
RESULTS
Patient characteristics
Our study involved 29 children and young adults with types III and IV OI, 5 children with type VI OI, 2 children with type VII OI and one young adult with type XIV OI. Median age was 23.6 years (range 3.1–49.6), with types III, IV, VI, VII or XIV OI median age was 26.7 (range 19.6–49.6), 26.2 (range 3.1–42.5), 12.7 (range 4.5–31.7), 7.0 (range 3.9–10.1) or 32.9 years, respectively (table 1). Sixteen of 37 individuals with OI were male; 21 were female. Genotyping identified COL1A1 variants in 17 individuals (7 with type III OI, 10 with type IV OI) or COL1A2 variants in 12 individuals (1 with type III OI, 11 with type IV OI). Questionnaire responses for behaviours potentially affecting respiration revealed that 2 of 8 individuals with type III and 3 of 21 with type IV OI smoked tobacco. Histories of significant respiratory infections were elicited from both children with type VII OI and 3 of 8 (37%), 4 of 21 (19%) and 1 of 5 (20%) individuals with types III, IV or VI OI, respectively. Chronic respiratory disease was self-reported in 2 of 8 (25%), 4 of 21 (19%), 1 of 5 (20%) and 1 of 2 (50%) of individuals with types III, IV, VI or VII OI, respectively. One of 8 (12.5%) individuals with type III and 9 of 21 (43%) type IV OI experienced chronic cough. Dyspnoea with activity was reported by 3 of 8 (37%), 7 of 21 (33%) and 1 of 5 (20%) individuals with types III, IV or VI OI, respectively. Walking aids were used by 7 of 8 (87%), 10 of 21 (48%) and 2 of 5 (40%) of individuals with types III, IV or VI OI, respectively, both children with type VII, and one adult with type XIV OI. Overall, more than one respiratory behaviour or symptom was reported by six individuals (one with type III OI and five with type IV). Five of these six individuals had a COL1A1 variant, and one had a COL1A2 variant.
Table 1.
Clinical characteristics and pulmonary function measurements of subjects with OI
| III OI (n=8) | IV OI (n=21) | VI OI (n=5) | VII OI (n=2) | XIV OI (n=1) | P value* | |
|---|---|---|---|---|---|---|
| Age: median (range) | 26.7 (19.6–49.6) | 26.2 (3.1–42.5) | 12.7 (4.5–31.7) | 7.0 (3.9–10.1) | 32.9 | |
| Male/Female | 3/5 | 10/11 | 2/3 | 0/2 | 1/0 | |
| COL1A1/COL1A2 variant | 7/1 | 10/11 | NA | NA | NA | |
| Smoking history, n (%) | 2 (25) | 3 (14) | 0 | 0 | 0 | |
| Prior respiratory infection, n (%) | 3 (38) | 4 (19) | 1 (20) | 2 (100) | 0 | |
| Chronic respiratory disease, n (%) | 2 (25) | 4 (19) | 1 (20) | 1 (50) | 0 | |
| Cough, n (%) | 1 (13) | 9 (43) | 0 | 0 | 0 | |
| Use of walking aid, n (%) | 7 (88) | 10 (48) | 2 (40) | 2 (100) | 1 (100) | |
| Dyspnoea with activity, n (%) | 3 (38) | 7 (33) | 1 (20) | 0 | 0 | |
| Dyspnoea at rest, n (%) | 1 (13) | 3 (14) | 0 | 0 | 0 | |
| Restrictive lung disease, n (%) | 7 of 7 (100) | 11 of 19 (58) | 1 of 5 (20) | 1 of 1 (100) | 0 | |
| Obstructive lung disease, n (%) | 1 of 7 (14) | 0 | 1 of 5 (20) | 0 | 0 | |
| FEV1%±SD | 35.3±14.5 | 76.0±19.3 | 81.8±18.2 | 56† | 103 | <0.001, 0.006, NS |
| FVC%±SD | 37.9±13.8 | 75.4±20.1 | 86±11.6 | 69† | 98 | <0.001, 0.002, NS |
| FEV1/FVC±SD | 0.80±0.06 | 0.86±0.06 | 0.84±0.2 | 0.91† | 0.85 | NS, NS, NS |
| FEF25%–75%±SD | 31.4±17.5 | 87.1±28.7 | 73.3±41.7 | NA | 112 | <0.001, 0.041, NS |
| TLC%±SD | 41.0±13.4 | 76.1±19.3 | 137±74.6 | NA | 98 | <0.001, 0.007, NS |
| DLCOa%±SD | 33.1±14.2 | 58.4±15.7 | 59.8±16.6 | NA | 73 | 0.001, 0.020, NS |
Chronic respiratory disease, included diagnoses of restrictive lung disease, asthma, sleep apnoea with or without use of CPAP. Sleep apnoea in OI is generally secondary to soft connective tissue in the upper airway and can be a significant respiratory system problem for these patients. Although sleep apnoea is an upper airway condition, it has a complex bidirectional relationship with chronic lung disease.33 Prior respiratory infections, included pneumonia with or without recurrence, bronchitis with or without recurrence, influenza, Scarlet fever with respiratory component.
P values of type III OI versus type IV OI, type III OI versus type VI OI, type IV OI versus type VI OI.
Data available from one individual with type VII OI.
CPAP, continuous positive airway pressure; DLCOa%, diffusion capacity percent predicted; FEF25%–75%, forced expiratory flow at 25%–75% of forced vital capacity percent predicted; FEV1%, forced expiratory volume in 1 s percent predicted; FVC, forced vital capacity; FVC%, forced vital capacity percent predicted; NA, data not available or not applicable; NS, not significant (ie, p>0.05); OI, osteogenesis imperfecta; TLC%, total lung capacity percent predicted.
Height surrogate comparison for PFT calculation in individuals with OI
We compared percentages of predicted PFTs calculated using both arm span and ulnar length as height surrogates. No significant differences in percentages of predicted FEV1, FEF25%–75%, TLC or DLCOa were found using either height surrogate in individuals with type III, IV or VI OI (figure 1A). Only FVC percent predicted values calculated using arm span measurements as surrogates for height were significantly higher than those calculated using ulnar length in individuals with type VI OI.
Figure 1.
Pulmonary function values in individuals with type III, IV or VI osteogenesis imperfecta (OI). (A) Comparison of pulmonary function test (PFT) results for forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), forced expiratory flow at 25%–75% of FVC (FEF25%–75%), total lung capacity (TLC) or diffusion capacity (DLCOa) in individuals with type III, IV or VI calculated using arm span (AS) or ulnar length (UL) as surrogate measurements for height. No significant differences were found, except individuals with type VI OI had higher FVC calculated using AS versus UL as surrogate for height. (B) Percentages of predicted FEV1, FVC, FEF25%–75%, TLC or DLCOa values calculated using AS as surrogates for height were significantly lower in type III OI compared with either type IV or VI OI. (C) In individuals with type III or type IV OI, percentage of predicted FEF25%–75% values calculated using AS as surrogates for height were significantly lower in those with COL1A1 variants compared with those with COL1A2 variants. FEV1, FVC, TLC and DLCOa normal range: 80%–120% predicted; FEF25%–75% normal range: ≥65% predicted.
PFT results in types III, IV and VI OI
Since arm span and ulnar length measurements yielded comparable results as height surrogates, PFTs calculated using arm span were analysed in this cohort with OI. PFTs showed restrictive lung disease, indicated by TLC <80% predicted, in more than half of these young patients, including 7 of 7 (100%), 11 of 19 (58%), 1 of 5 (20%) and 1 of 1 individuals with type III, IV, VI or VII OI, respectively. Obstructive lung disease, indicated by FEV1/FVC ratio <70%, was found in two individuals, one with type III OI and one with type VI OI (table 1). It is notable that gas exchange or diffusion capacity was abnormal in 90% of all individuals with OI.
Among individuals with type IV OI, mean age and use of walking aids of patients with restrictive lung disease (31.6 years and 8 of 11 individuals, respectively) were significantly higher than those without restrictive lung disease (21.0 years and 1 of 8 individuals, respectively) (table 2). In addition, mean thoracic Cobb angle tended to be greater in patients with type IV OI with restrictive lung disease (28.3°) compared with those without restrictive lung disease (11.1°) (p=0.06).
Table 2.
Clinical characteristics of subjects with type IV OI with and without restrictive lung disease
| Type IV OI with restrictive lung disease (n=11) | Type IV OI without restrictive lung disease (n=8) | P value | |
|---|---|---|---|
| Age: median (range) | 32.2 (14.5–42.5) | 20.9 (9.4–32.4) | |
| Male/Female | 6/5 | 4/4 | NS |
| COL1A1/COL1A2 variant | 5/6 | 5/3 | NS |
| Smoking history, n (%) | 2 (18) | 1 (13) | NS |
| Prior respiratory infection, n (%) | 4 (36) | 0 | NS |
| Chronic respiratory disease, n (%) | 3 (27) | 0 | NS |
| Cough, n (%) | 6 (56) | 1 (13) | NS |
| Use of walking aid, n (%) | 8 (73) | 1 (13) | 0.02 |
| Dyspnoea, n (%) | 5 (45) | 2 (25) | NS |
| Cobb angle, degrees | 28.3±20.9 | 11.1±9.7 | 0.06 |
| Thoracic bone deformities or fractures, n (%) | 7 (64) | 2 (25) | NS |
| Vertebral rods, n (%) | 3 (27) | 1 (13) | NS |
NS, not significant (ie, p>0.05); OI, osteogenesis imperfecta.
PFT results correlate with severity of OI types
PFT results were compared in three OI types. Percentages of predicted FEV1, FVC, FEF25%–75%, TLC or DLCOa in individuals with type III OI were significantly lower than in individuals with type IV or type VI OI (figure 1B) (table 1). No significant differences were found between individuals with type IV or type VI OI.
Aggregate data showed that individuals with type III OI had small airways airflow obstruction (ie, FEF25%–75% <65% predicted), severe restriction and severe reduction of diffusion capacity. Individuals with type IV OI had mild restriction and moderate reduction of diffusion capacity, and individuals with type VI OI had hyperinflation and moderate reduction of diffusion capacity. One child with type VII OI had spirometry values suggestive of restriction. One individual with type XIV OI had mild reduction of diffusion capacity.
PFT results in individuals with COL1 genetic variants
We compared PFTs in individuals with COL1A1 vs COL1A2 pathological variants. We found that FEF25%–75% was significantly lower in individuals with COL1A1 variants compared with those with COL1A2 variants, whose mean FEF25%–75% was normal (figure 1C). In contrast, mean FEF25%–75% measurement in individuals with COL1A1 variants was 60.6% predicted, indicating that this group of patients with OI had small airways airflow obstruction.
Pulmonary and extrapulmonary structure in individuals with OI
Chest CT scans identified structural abnormalities of the airways, lung parenchyma or extrapulmonary tissue in OI (figure 2) (table 3). The most common pulmonary finding was small bronchial thickening at the level of subsegmental bronchi. Abnormal bronchial thickening was defined as a ratio of airway wall thickness to airway diameter of 0.20 or greater.21 22 Small bronchial wall thickening was found in all individuals, except 3 of 21 individuals with type IV OI. These three individuals had COL1A2 variants; two of these three individuals had no detectable thoracic bone disease. The small bronchi ratio was significantly higher in individuals with type VI OI than those with type IV OI (p<0.001). No significant differences in trachea or large bronchi ratios were found between groups. Bronchiectasis was found in 3 of 8 (37.5%) and 2 of 21 (9.5%) of individuals with type III or IV OI, respectively.
Figure 2.
Chest CT scan findings in individuals with osteogenesis imperfecta (OI). Representative chest CT scan images are shown. (A) Supine CT scan image demonstrates focal emphysema (open black arrow) in the left lower lobe, ground glass opacities (black circles) and bony chest wall deformities in an individual with type III OI. (B) Prone high-resolution CT (HRCT) scan image shows bilateral lower lobe bronchiectasis (open white arrows) in an individual with type IV OI without significant chest wall defects. (C) Bilateral lower lobe bronchiectasis is detected in a prone HRCT scan image from an individual with type IV OI with chest wall deformities, gracile ribs and severe rotary kyphoscoliosis. (D and E) A small right upper lobe lung cyst (solid white arrowhead) and right lower lobe atelectasis (solid black arrow) are found in prone HRCT scan images from a child with type VII OI. (F) Diffuse low lung tissue attenuation in bilateral lobes is observed in a prone HRCT scan image from an individual with type XIV OI. (G and H) Pleural thickening (solid white arrow) is found in supine CT scan images from individuals with type III (G) or type IV OI (H). (I) Small bronchial airway thickening (white circle) is detected in an individual with type IV OI.
Table 3.
Chest CT scan findings of subjects with OI
| III OI (n=8) | IV OI (n=21) | VI OI (n=5) | VII OI (n=2) | XIV OI (n=1) | P value* | |
|---|---|---|---|---|---|---|
| Bronchial thickening, n (%) | 8 (100) | 18 (86) | 5 (100) | 2 (100) | 1 (100) | |
| Small bronchi ratio±SD | 0.27±0.06 | 0.25±0.04 | 0.33±0.03 | 0.28±0.01 | 0.25 | NS, NS, <0.001 |
| Large bronchi ratio±SD | 0.12±0.02 | 0.11±0.02 | 0.13±0.02 | 0.15±0.0 | 0.09 | NS, NS, NS |
| Trachea ratio±SD | 0.10±0.02 | 0.10±0.02 | 0.12±0.02 | 0.13±0.02 | 0.07 | NS, NS, NS |
| Bronchiectasis, n (%) | 3 (38) | 2 (10) | 0 | 0 | 0 | |
| Atelectasis, n (%) | 7 (88) | 9 (43) | 2 (40) | 2 (100) | 0 | |
| Reticulations, n (%) | 4 (50) | 6 (29) | 1 (20) | 1 (50) | 0 | |
| Ground glass, n (%) | 6 (75) | 1 (5) | 0 | 2 (100) | 0 | |
| Emphysema, n (%) | 1 (13) | 4 (19) | 1 (20) | 0 | 1 (100) | |
| Focal lung tissue density ≤ −910 Hounsfield units, n (%) | 4 (50) | 12 (57) | 2 (40) | 0 | 1 (100) | |
| Whole lung tissue density±SD, Hounsfield units | −846±46.9 | −843±61.3 | −838±54.9 | −763±56.3 | −912±17.8 | NS, 0.045, NS |
| Cysts, n (%) | 0 | 0 | 0 | 1 (50) | 0 | |
| Pleural thickening, n (%) | 5 (63) | 10 (48) | 1 (20) | 2 (100) | 0 | |
| Rib deformities, n (%) | 6 (75) | 8 (38) | 1 (20) | 1 (50) | 0 | |
| Rib fractures, n (%) | 6 (75) | 4 (19) | 5 (100) | 2 (100) | 0 | |
| Vertebral deformities, n (%) | 6 (75) | 4 (19) | 3 (60) | 1 (50) | 0 | |
| Vertebral fractures, n (%) | 5 (63) | 5 (24) | 3 (60) | 2 (100) | 0 | |
| Scoliosis, n (%) | 7 (88) | 15 (71) | 1 (20) | 2 (100) | 1 (100) | |
| Vertebral rods, n (%) | 5 (63) | 4 (19) | 1 (20) | 0 | 0 |
P values of type III OI versus type IV OI, type III OI versus type VI OI, type IV OI versus type VI OI.
NS, not significant (ie, p>0.05); OI, osteogenesis imperfecta.
Abnormal findings affecting lung parenchyma
Atelectasis, reticulation, ground glass opacities and emphysema were detected in this OI cohort. Atelectasis was found in many cases, including 7 of 8 (87.5%), 9 of 21 (43%), 2 of 5 (40%) individuals with type III, IV or VI, respectively, as well as both individuals with type VII OI. Reticulations were detected in 4 of 8 (50%), 6 of 21 (29%), 1 of 5 (20%) and 1 of 2 individuals with type III, IV, VI or VII OI, respectively. Ground glass opacities were found in 6 of 8 (75%) and 1 of 5 (20%) individuals with type III or IV OI, respectively, and both individuals with type VII OI.
Regions of emphysema were visualised in 1 of 8 (12.5%), 4 of 21 (19%) and 1 of 5 (20%) individuals with type III, IV or VI OI, respectively, and in one adult with type XIV OI. A threshold of −910 Hounsfield units was used to identify emphysema.23 24 At least one pixel consistent with focal emphysema was identified in 4 of 8 (50%), 12 of 21 (57%), 2 of 5 (40%) individuals with type III, IV or VI OI, respectively, and one patient with type XIV OI. Mean Hounsfield unit measurements were consistent with emphysema in one lobe in 2 of 21 (9.5%) individuals with type IV OI and in three lobes in the individual with type XIV OI, whose whole lung tissue density was consistent with emphysema (data not shown). Whole lung tissue density in type III OI was significantly lower than in those with type VI OI (p=0.045). Furthermore, whole lung tissue density in individuals with COL1A2 variants was significantly lower than in those with COL1A1 variants (p<0.001) (online supplemental figure 1). A small lung cyst was detected in 1 of 2 children with type VII OI. Pulmonary infiltrates or abnormal lung nodules were not found.
In extrapulmonary tissue, small areas of pleural thickening were found in 5 of 8 (62.5%), 10 of 21 (47.6%), 1 of 5 (20%) individuals with type III, IV or VI OI, respectively, and both children with type VII OI. Pleural effusions were not identified. Multiple rib or vertebral deformities or fractures were detected in type III, IV, VI or VII OI; no bony deformities or fractures were found in the adult with type XIV OI. Scoliosis was found in 26 of 37 (70%) individuals with OI, including 7 of 8 (87.5%), 15 of 21 (71%), 1 of 5 (20%) individuals with type III, IV or VI OI, respectively, both individuals with type VII OI, and the one adult with type XIV OI. Surgical rods were present in 5 of 8 (62.5%), 4 of 21 (19%) and 1 of 5 (20%) individuals with type III, IV or VI OI, respectively.
Pulmonary function correlations in individuals with OI
Our data on this patient cohort also supported the current understanding correlating decline in pulmonary function with severity of scoliosis. We found that percentages of predicted FEV1, FVC, FEF25%–75% or DLCOa correlated negatively with Cobb angle in all individuals with OI (figure 3). Given the wide age range of 46 years in this cohort, we explored correlations between lung function and age. Percentages of predicted FEV1, FVC or TLC correlated negatively with age in all individuals with OI.
Figure 3.
Correlation between scoliosis or age and pulmonary function in individuals with osteogenesis imperfecta (OI). Forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), forced expiratory flow at 25%–75% of FVC (FEF25%–75%) and diffusion capacity (DLCOa) percentage of predicted values correlated inversely with thoracic Cobb angle in all individuals with OI. FEV1, FVC and total lung capacity (TLC) percentage of predicted values correlated inversely with age in all individuals with OI.
DISCUSSION
In the present study, we extended our prior examination of intrinsic lung disease in OI9 to 37 children and young adults with OI, most of whom have classical OI with collagen structural abnormalities, and who underwent chest CT and PFTs. Because of the significant and disproportionate short stature of OI, calculation of PFTs using patient height is inaccurate.25 Our data demonstrate that using arm span or ulnar length measurements as surrogates for height yield comparable results in patients with OI.
In the PFTs of our OI cohort, TLC is reduced below 80% predicted, indicative of restrictive disease, in all patients with type III OI and about half of patients with moderate type IV OI. Notably, mean TLC in type III OI is already markedly reduced to 40% of predicted at this relatively young age. Within type IV OI, those without restrictive disease were younger and did not use walking aids compared with those with restrictive disease. Thus, restrictive lung disease generally corresponds with severity of OI, with restrictive disease more severe in type III than in those type IV OI who are older and less mobile, who in turn have greater severity than the type IV who are younger and more mobile. Illustrating involvement of both intrinsic and extrinsic factors in OI lung pathology, those with restrictive disease tended to have greater mean Cobb angle to their scoliosis. Furthermore, it was striking that diffusion capacity for carbon monoxide was reduced in almost all classical patients with OI, again with type III significantly worse than type IV OI. This factor correlates with diffuse parenchymal lung disease and corroborates the CT scan findings. A recent study of 30 patients with OI with an average age of 39 years did not detect decreased diffusion capacity.22 The reason for this difference between studies is not clear, although the technical challenges of measuring diffusion capacity must be considered (see ‘Materials and methods’ section), as well as differences in the genetic aetiology of OI in this cohort, which included 14 individuals whose causative mutation had not been determined, and 9 individuals with type I OI. Type I OI is a mild form caused by quantitative collagen deficiency rather than collagen structural abnormalities. The NIH cohort did not include individuals with type I OI.
For air flow, PFTs reveal significant reduction in FEV1, FVC and FEF25%–75% in individuals with type III OI. However, the FEV1/FVC ratio, which detects obstructive lung disease involving large airways, is abnormal in only two individuals. FEV1 decreased with age, whether measured by regression curve or by decades, so further longitudinal examination of air flow is warranted to determine whether progression to obstructive disease is part of the phenotype.
Comprehensive review of CT scans from 37 individuals with OI reveals a broad spectrum of abnormal pulmonary structure. We localised highly prevalent airway wall thickening to the small airways at the level of subsegmental bronchi in type III and IV OI. Another recent study also reported bronchial wall thickening as a common finding.22 We also show that individuals with type VI, VII or XIV OI, which are uncommon OI types, have bronchial wall thickening. While it remains unclear whether bronchial wall thickening is associated with altered airflow in OI, we found small airway obstruction only in type III OI. Furthermore, FEF25%–75%, a measure of air flow in small airways, was the only PFT component that was significantly worse in individuals with COL1A1 vs COL1A2 variants, again indicating small airway flow is a component of a more severe OI respiratory phenotype. Bronchial wall thickening in OI may be directly related to abnormal collagen or a secondary inflammatory response. Measurement of fractional exhaled nitric oxide, a non-invasive biomarker of airway inflammation, may elucidate the potential role of airway inflammation in OI. Although Fractional Exhaled Nitric Oxide (FeNO) is commonly used to measure allergen-induced airway inflammation, FeNO can also measure airway inflammation of other aetiologies,26 and thus may be useful in elucidating the pathophysiology of bronchial wall thickening in OI.
Our study also independently validates a prior report of ground glass opacities in patients with OI.22 We identified ground glass infiltrates in most individuals with type III OI, one with type IV OI and both with type VII OI. The aetiology of ground glass opacities in OI is unknown, but data from the Aga2 mouse, an animal model of type III OI with a Col1a1 defect, suggest that lung haemorrhage or alveolar infiltration by neutrophils and macrophages may be contributing factors.9 These lung findings in Aga2 mice suggest that vascular fragility or defects in angiogenesis are features of these OI types. Further studies in animal models of OI are indicated to determine underlying mechanisms of lung disease in OI.
Interstitial reticulations and regions of emphysema were identified in some of our patients with OI. This dichotomy demonstrates that collagen-related disorders are associated with a wide phenotypic spectrum in alveolar tissue, ranging from deficient (emphysema) to excessive (interstitial lung disease/fibrosis) levels of collagen. Emphysema was reported in Col1a1Jrt/+ and Crtap−/− mice, which are murine models of OI associated with collagen structural and post-translational modification defects, respectively.10 11 27 Inhibition of transforming growth factor (TGF)-β in Crtap−/− mice ameliorated pulmonary disease, suggesting that TGF-β is involved in the pathobiology of emphysema in some types of recessive OI.28 Only limited information is available regarding emphysema in patients with OI. Emphysema was reported in a patient with type III OI, and histological findings of emphysema were reported in newborns with type II OI and a patient with OI aged 19 years.6 22 29 In our study, emphysematous regions were visualised in CT scans of 7 patients with OI, and low lung tissue density measurements consistent with emphysema were found in 19 patients with OI. Notably, an adult non-smoker with type XIV OI with radiographic evidence of diffuse emphysema had corresponding impairment of diffusion capacity. Type XIV OI is an uncommon autosomal recessive OI type associated with variants in TMEM38B, which encodes trimeric intracellular cation channel type B, a ubiquitously expressed ER-membrane channel involved with intracellular calcium release.30 Although few patients with type XIV OI were reported, an individual with type XIV OI aged 22 years developed lower airway obstruction and air trapping, which could be consistent with emphysema.31 Furthermore, meta-analysis of genetic variants associated with chronic obstructive pulmonary disease identified TMEM38B as a leading gene.32 Further studies are indicated to expand the understanding of emphysema in OI and to explore the potential pathobiological role of TMEM38B.
We identified pleural thickening in almost half of our cohort with OI. A paucity of information is available regarding pleural involvement in OI, with only one reported case of pleural thickening in a histopathological specimen from a patient with OI aged 19 years.29 Lung reticulations and pleural thickening in OI may be consequences of abnormal fibrotic processes in alveolar interstitial and extrapulmonary tissue. Excessive TGF-β signalling was shown to contribute to bone disease and emphysema in the Crtap−/− murine model of OI and in the Pigment Epithelium-Derived Factor (PEDF) knockout mouse for type VI OI.15 28 Given the profibrotic effects of TGF-β, it is possible that the TGF-β pathway may also be involved in development of interstitial lung disease and pleural thickening in OI. Alternatively, pleural thickening in OI could be secondary to residual effects of prior respiratory infection or chest wall injury.
Overall, these findings expand the phenotype of intrinsic disease involving pulmonary parenchyma, lung airways and pleura in OI. While atelectasis can also be found in patients with scoliosis of other origins, bronchiectasis, reticulations, ground glass opacities, emphysema and bronchial thickening would not be expected to be secondary to scoliosis. Respiratory disease is a prominent cause of mortality in OI, and studies to identify therapeutic targets could lead to development of effective treatment for this aspect of OI.
Supplementary Material
WHAT IS ALREADY KNOWN ON THIS TOPIC
Osteogenesis imperfecta (OI) is a heterogeneous group of inherited bone dysplasias characterised by skeletal and extraskeletal manifestations.
Respiratory disease is the leading cause of morbidity and mortality in patients with OI.
Despite previous thought that lung disease in OI was secondary to scoliosis and other skeletal deformities, research in a cohort of young children with OI showed significant decline in pulmonary function with age, independent of scoliosis.
Genetically modified mice with collagen type 1 defects also display cardiopulmonary dysfunction independent of skeletal abnormalities.
WHAT THIS STUDY ADDS
Our findings expand the phenotype of respiratory disease in OI to young adults, demonstrating restrictive disease and impaired gas exchange in most patients independent of skeletal abnormalities, such as scoliosis, using pulmonary function tests (PFTs).
Analysis of chest computed tomography (CCT) and PFT—which showed highly prevalent small airway bronchial thickening and decreased FEF25%–75%, indicating reduced airflow through the small airways, respectively—indicate a role of small airway abnormalities in the pathobiology of OI respiratory disease.
We also provide the first demonstration that arm span and ulnar length measurements are comparable height surrogates for calculating and interpreting PFT results in patients with OI.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The findings of this study provide critical insight into origin, progression and manifestations of pulmonary disease in OI that may lead to improved treatment options and therapies for patients before they impact quality of life.
Acknowledgements
We thank our patients for participating in our research studies.
Funding
This work is supported by the Intramural Research Programme of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (ZIA HD008830-15 and ZIA HD000408-38), the National Human Genome Research Institute and the National Heart, Lung, and Blood Institute.
Footnotes
Competing interests None declared.
Ethics approval This study was approved by National Institutes of Health (NIH) institutional review board, protocol 18-CH-0120. Participants gave informed consent to participate in the study before taking part.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
Data availability statement
Data are available on reasonable request. Full data available on reasonable request.
REFERENCES
- 1.Marini JC, Forlino A, Bächinger HP, et al. Osteogenesis imperfecta. Nat Rev Dis Primers 2017;3:17052. [DOI] [PubMed] [Google Scholar]
- 2.Sillence DO, Senn A, Danks DM. Genetic heterogeneity in osteogenesis imperfecta. J Med Genet 1979;16:101–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Forlino A, Marini JC. Osteogenesis imperfecta. Lancet 2016;387:1657–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.McAllion SJ, Paterson CR. Causes of death in osteogenesis imperfecta. J Clin Pathol 1996;49:627–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Folkestad L, Hald JD, Canudas-Romo V, et al. Mortality and causes of death in patients with osteogenesis imperfecta: a register-based nationwide cohort study. J Bone Miner Res 2016;31:2159–66. [DOI] [PubMed] [Google Scholar]
- 6.Storoni S, Treurniet S, Micha D, et al. Pathophysiology of respiratory failure in patients with osteogenesis imperfecta: a systematic review. Ann Med 2021;53:1676–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yonko EA, Emanuel JS, Carter EM, et al. Respiratory impairment impacts QOL in osteogenesis imperfecta independent of skeletal abnormalities. Arch Osteoporos 2020;15:153. [DOI] [PubMed] [Google Scholar]
- 8.Widmann RF, Bitan FD, Laplaza FJ, et al. Spinal deformity, pulmonary compromise, and quality of life in osteogenesis imperfecta. Spine 1999;24:1673. [DOI] [PubMed] [Google Scholar]
- 9.Thiele F, Cohrs CM, Flor A, et al. Cardiopulmonary dysfunction in the osteogenesis imperfecta mouse model Aga2 and human patients are caused by bone-independent mechanisms. Hum Mol Genet 2012;21:3535–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Baldridge D, Lennington J, Weis M, et al. Generalized connective tissue disease in Crtap−/−mouse. PLoS One 2010;5:e10560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dimori M, Heard-Lipsmeyer ME, Byrum SD, et al. Respiratory defects in the CrtapKO mouse model of osteogenesis imperfecta. Am J Physiol Lung Cell Mol Physiol 2020;318:L592–605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Tsukui T, Sun K-H, Wetter JB, et al. Collagen-producing lung cell atlas identifies multiple subsets with distinct localization and relevance to fibrosis. Nat Commun 2020;11:1920. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Marini JC, Dang DA AN. Osteogenesis imperfecta. In: Feingold KR, Anawalt B, Boyce A, eds. Endotext. South Dartmouth (MA): MDText. com, Inc, 2000. Available: https://www.ncbi.nlm.nih.gov/books/NBK279109/ [Google Scholar]
- 14.Garibaldi N, Besio R, Dalgleish R, et al. Dissecting the phenotypic variability of osteogenesis imperfecta. Dis Model Mech 2022;15:dmm049398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kang H, Aryal ACS, Marini JC. Osteogenesis imperfecta: new genes reveal novel mechanisms in bone dysplasia. Transl Res 2017;181:27–48. [DOI] [PubMed] [Google Scholar]
- 16.Jovanovic M, Guterman-Ram G, Marini JC. Osteogenesis imperfecta: mechanisms and signaling pathways connecting classical and rare OI types. Endocr Rev 2022;43:61–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ballenger KL, Tugarinov N, Talvacchio SK, et al. Osteogenesis imperfecta: the impact of genotype and clinical phenotype on adiposity and resting energy expenditure. J Clin Endocrinol Metab 2022;107:67–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Graham BL, Steenbruggen I, Miller MR, et al. Standardization of spirometry 2019 update. An official American thoracic society and European respiratory society technical statement. Am J Respir Crit Care Med 2019;200:e70–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.El-Chemaly S, Cheung F, Kotliarov Y, et al. The immunome in two inherited forms of pulmonary fibrosis. Front Immunol 2018;9:76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.O’Brien KJ, Introne WJ, Akal O, et al. Prolonged treatment with open-label pirfenidone in Hermansky-Pudlak syndrome pulmonary fibrosis. Mol Genet Metab 2018;125:168–73. [DOI] [PubMed] [Google Scholar]
- 21.Matsuoka S, Uchiyama K, Shima H, et al. Bronchoarterial ratio and bronchial wall thickness on high-resolution CT in asymptomatic subjects: correlation with age and smoking. AJR Am J Roentgenol 2003;180:513–8. [DOI] [PubMed] [Google Scholar]
- 22.Khan SI, Yonko EA, Carter EM, et al. Cardiopulmonary status in adults with osteogenesis imperfecta: intrinsic lung disease may contribute more than scoliosis. Clin Orthop Relat Res 2020;478:2833–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kinsella M, Müller NL, Abboud RT, et al. Quantitation of emphysema by computed tomography using a “density mask” program and correlation with pulmonary function tests. Chest 1990;97:315–21. [DOI] [PubMed] [Google Scholar]
- 24.Coxson HO, Rogers RM, Whittall KP, et al. A quantification of the lung surface area in emphysema using computed tomography. Am J Respir Crit Care Med 1999;159:851–6. [DOI] [PubMed] [Google Scholar]
- 25.Tam A, Chen S, Schauer E, et al. A multicenter study to evaluate pulmonary function in osteogenesis imperfecta. Clin Genet 2018;94:502–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Menzies-Gow A, Mansur AH, Brightling CE. Clinical utility of fractional exhaled nitric oxide in severe asthma management. Eur Respir J 2020;55:1901633. [DOI] [PubMed] [Google Scholar]
- 27.Baglole CJ, Liang F, Traboulsi H, et al. Pulmonary and diaphragmatic pathology in collagen type I α1 mutant mice with osteogenesis imperfecta. Pediatr Res 2018;83:1165–71. [DOI] [PubMed] [Google Scholar]
- 28.Grafe I, Yang T, Alexander S, et al. Excessive transforming growth factor-β signaling is a common mechanism in osteogenesis imperfecta. Nat Med 2014;20:670–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Morikawa M, Fukuda Y, Terasaki Y, et al. Osteogenesis imperfecta associated with dendriform pulmonary ossification. Am J Respir Crit Care Med 2016;193:460–1. [DOI] [PubMed] [Google Scholar]
- 30.Shaheen R, Alazami AM, Alshammari MJ, et al. Study of autosomal recessive osteogenesis imperfecta in Arabia reveals a novel locus defined by TMEM38B mutation. J Med Genet 2012;49:630–5. [DOI] [PubMed] [Google Scholar]
- 31.Webb EA, Balasubramanian M, Fratzl-Zelman N, et al. Phenotypic spectrum in osteogenesis imperfecta due to mutations in TMEM38B: unraveling a complex cellular defect. J Clin Endocrinol Metab 2017;102:2019–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Wyss AB, Sofer T, Lee MK, et al. Multiethnic meta-analysis identifies ancestry-specific and cross-ancestry loci for pulmonary function. Nat Commun 2018;9:2976. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Locke BW, Lee JJ, Sundar KM. OSA and chronic respiratory disease: mechanisms and epidemiology. Int J Environ Res Public Health 2022;19:5473. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available on reasonable request. Full data available on reasonable request.