Abstract
Osteogenesis imperfecta (OI) is an inherited disorder characterized by bone fragility with extraskeletal manifestations mostly due to COL1A1 and COL1A2 variants. Currently, 23 genes have been implicated in the pathogenesis of OI; however, literature on genotype–phenotype correlation and incidence of non-skeletal clinical features are limited. This study aims to identify genotype–phenotype correlations in patients with OI, allowing clinicians to better inform families of prognosis, optimize patient care, and facilitate evidence-based clinical decision-making. We retrospectively reviewed 294 patients with OI to collect demographic data, clinical characteristics, and genotypic information. Patients were stratified by COL1A1/1A2 vs non-COL1A1/1A2 variants to evaluate differences in phenotype. The majority of OI was due to variants in COL1A1/1A2 (91%), with the remaining 9% due to non-COL1A1/1A2 variants. Most patients in the COL1A1/2 group were White compared to the non-COL1A1/2 group (78% vs 50%; p = 0.004). COL1A/1A2 patients had higher incidence of blue sclerae (83% vs 58%, p = 0.002), dentinogenesis imperfecta (49% vs 15%, p < 0.001), and family history of OI (34% vs 12%, p = 0.03). Those in the non-COL1A1/1A2 group have higher rates of scoliosis compared to those in the COL1A1/1A2 group (62% vs 40%, p = 0.04), as well as higher rates of expressive language disorder/delay (15% vs 0.4% in non-COL1A1/1A2 and COL1A1/1A2 patients, respectively; p < 0.001). Identifying the underlying molecular etiology early is imperative for optimal clinical care, allowing for appropriate risk counseling, identification of affected relatives, and improved anticipatory care and management. These data support that rare subtypes of OI occur more frequently in non-White individuals and demonstrated genetic associations with incidence of blue sclera, dentinogenesis imperfecta, scoliosis, and expressive language disorders.
Keywords: genotype–phenotype, OI, pediatrics, orthopedics, genetic variants
Introduction
Osteogenesis imperfecta (OI) is a heterogenous connective tissue disorder with an estimated incidence of 1 per 10 000 individuals.1 The core phenotype of OI is bone fragility with recurrent fractures, skeletal deformity, and short stature; however, multiple secondary features have been described. OI is caused by dominant, recessive, or X linked variants in multiple genes resulting in variable phenotypes affecting both skeletal and extraskeletal tissues.1 The majority of OI is due to autosomal dominant variants in COL1A1 or COL1A2, the genes that code for the alpha chains of type I collagen, the most abundant protein in the bone matrix which is integral to its structure. People with COL1A1 and COL1A2 variants typically either have a quantitative changes resulting in decreased amounts of normal type I collagen, mainly the result of frameshift or nonsense mutations.2 Alternatively, those with more severe forms of the disease tend to have qualitative changes, typically the result amino acid substitutions in polypeptide chains, resulting in structurally abnormal type I collagen.3
Autosomal dominant variants in these genes were first associated with OI in the late 1970s.4 Classification of OI types I-IV was proposed by Sillence et al. in 1979 based on clinical and radiographic features and inheritance patterns, which has been the standard since that time. Types II and III are the most severe, with type II typically resulting in perinatal demise, followed in severity by types IV and I. Individuals with type I typically have grossly normal stature but remain prone to frequent fractures.5 In 2015, a revision to the Nosology and Classification of Genetic Skeletal Disorders distinguished 5 OI types based on clinical presentation.6 As bioinformatics and sequencing technology has grown in use and accessibility, it has catalyzed the discovery of new autosomal recessive variants in non-COL1A1/A2 genes in OI, necessitating the Genetic classification of 18 types 1 yr later in 2016.7 The most recent nosology of genetic skeletal disorders published in 2023 has grouped skeletal disorders with one of the groups the OI and bone fragility group. The current recommendation for naming OI is to note the phenotypic severity and what gene is related to the disorder.8 Currently, the Online Mendelian Inheritance in Man (MIM) database lists new types of OI with the discovery of novel genes (https://www.ncbi.nlm.nih.gov/omim/).
Of non-COL1A1/A2 variants, genetic defects leading to compromised bone mineralization, defects in collagen post-translational modification, aberrant collagen processing and cross-linking, or altered osteoblast differentiation and function have been discovered in patients with OI.1 While this new information improves patient care and genetic counseling, the phenotypic and clinical heterogeneity of those with non-COL1A1/1A2 variants often complicate classification and prognosis. Additionally, the large number of candidate genes further complicates the identification of consistent genotype–phenotype presentations.
Many have studied genotype–phenotype correlations in OI2,9–18; however, these were conducted with relatively small sample sizes at institutions outside the United States and typically have focused on those with COL1A1/1A2 inherited defects. The present study aims to augment current literature with one of the largest cohorts evaluated and further elucidate the complex relationships and phenotypic differences between those with COL1A1/1A2 and non-COL1A1/1A2 variants.
Materials and methods
Study population
We retrospectively reviewed all patients with a diagnosis of OI (ICD code Q78.0) who were treated at Children’s Nebraska in Omaha, Nebraska between January 1990 and June 2022. A total of 447 patients were initially identified from the electronic health record and our institution’s OI Database; however, 107 patients were excluded as they did not have a clinical OI type documented. An additional 46 patients were excluded due to incomplete medical records resulting in a total of 294 patients included in the final analysis. Approval for the study was obtained from the University of Nebraska Medical Center Institutional Review Board (#0498-22-EP).
Patient demographics (sex, age in months, ethnicity, clinical OI type, and genetic variant) were collected for each patient. Clinical characteristics were collected and include the presence of blue sclera, dentinogenesis imperfecta (DI), hearing loss- and age at first hearing loss if applicable, bone mineral density, use of bisphosphonates, rod implantation for long bone deformity, scoliosis- and scoliosis operations if applicable, family history of OI, joint hypermobility, and if the patient has been diagnosed with short stature. This study also evaluated the prevalence of non-OI diagnoses including the prevalence of autism spectrum disorder (ASD), expressive language disorders/delays, seizures, tics, and hydrocephalus or necessity of ventriculoperitoneal shunt (VPS). Assessments occurred with annual multidisciplinary clinic evaluations including long bone and spinal radiographs, DEXA scans, physical and occupational therapy evaluations, and routine pulmonary and audiologic evaluation.
Demographics and clinical characteristics were compared for patients with COL1A1/1A2 variants to patients with non-COL1A1/1A2 variants. Patients were further stratified using a functional metabolic classification proposed by Forlino and Marini. This consists of 5 functional groups: group A consists of primary defects in collagen structure/processing (COL1A1, COL1A2, and BMP1), group B is due to defects in collagen modification (CRTAP, LEPRE1, PPIB, and TMEM38B), group C relates to defects in collagen folding and cross-linking (SERPINH1, FKBP10, and PLOD2), group D consists of defection in ossification/mineralization (IFITM5 and SERPINF1), and group E with defects in osteoblast development with collagen insufficiency (WNT1, CREB3L1, and SP7).18
Statistical analysis
Descriptive statistics which included means and standard deviations for all continuous variables and counts and percentages for categorical responses were used to summarize the data. Three new continuous variables were created: age at first hearing loss diagnosis, age at first bisphosphonate infusion, and age at DEXA scan. Age values were rounded down to the nearest whole integer. Chi square or Fisher’s Exact test were used as appropriate to compare all categorical variables by COL groups. Independent samples t-tests or Wilcoxon Rank Sum tests were used to compare continuous variables by COL groups. All analyses were conducted using SAS version 9.4. p < 0.05 was considered statistically significant.
Results
Study population
Of the 294 individuals included in this study, the majority had COL1A1/1A2 variants (n = 268, 91.2%). There was equal representation between males and females, overall, and no significant difference in gender representation between groups (p = 0.22). There was a significant association between race and COL1A1/1A2 grouping. The majority of patients in the COL1A1/1A2 group were White compared to the non-COL1A1/1A2 group (78% vs 50%; p = 0.002). Population characteristics are displayed stratified by COL vs non-COL grouping and by individual grouping in Tables 1 and 2, respectively.
Table 1.
Raw data and statistics stratified by COL vs non-COL mutation.
| COL groups | ||||
|---|---|---|---|---|
|
COL
(N = 268) |
Non-COL
(N = 26) |
Total
(N = 294) |
p-value | |
| Sex, n (%) | 0.22a | |||
| Female | 131 (48.9%) | 16 (61.5%) | 147 (50.0%) | |
| Male | 137 (51.1%) | 10 (38.5%) | 147 (50.0%) | |
| Race, n (%) | 0.004 b | |||
| White | 209 (78.0%) | 13 (50.0%) | 222 (75.5%) | |
| Non-White | 59 (22.0%) | 13 (50.0%) | 72 (24.5%) | |
| Autism, n (%) | 4 (1.5%) | 2 (7.7%) | 6 (2.0%) | 0.09b |
| Expressive language disorder/delay, n (%) | 1 (0.4%) | 4 (15.4%) | 5 (1.7%) | 0.0002 b |
| Blue sclera, n (%) | 221 (82.5%) | 15 (57.7%) | 236 (80.3%) | 0.002 a |
| Dentinogenesis imperfecta, n (%) | 131 (48.9%) | 4 (15.4%) | 135 (45.9%) | 0.0009 b |
| Age at first hearing loss diagnosis, mean (SD) | 8.6 (6.16) | 3.8 (2.99) | 8.2 (6.09) | 0.13c |
| Age at first bisphosphonate infusion, mean (SD) | 3.9 (4.52) | 2.5 (3.36) | 3.7 (4.44) | 0.39c |
| Age at DEXA, mean (SD) | 11.4 (5.10) | 8.9 (4.59) | 11.2 (5.10) | 0.02 d |
| Patient received bisphosphonates at time of DEXA, n (%) | 231 (87.5%) | 26 (100.0%) | 257 (88.6%) | 0.05b |
| DEXA Z-score, mean (SD) | −1.1 (1.73) | −0.9 (2.08) | −1.1 (1.76) | 0.68d |
| Family history of OI, n (%) | 91 (34.0%) | 3 (11.5%) | 94 (32.0%) | 0.03 b |
| History of scoliosis, n (%) | 108 (40.4%) | 16 (61.5%) | 124 (42.3%) | 0.04 a |
| Scoliosis operations, n (%) | 21 (19.4%) | 1 (6.3%) | 22 (17.7%) | 0.30b |
| History of rod placement for long bone deformity, n (%) | 184 (68.7%) | 20 (76.9%) | 204 (69.4%) | 0.38a |
| Presence of joint hypermobility, n (%) | 207 (77.2%) | 21 (80.8%) | 228 (77.6%) | 0.68a |
| Patient diagnosed short in stature, n (%) | 145 (54.1%) | 19 (73.1%) | 164 (55.8%) | 0.06a |
aChi-square p-value;
bFisher Exact p-value;
cWilcoxon rank sum p-value;
dEqual variance 2 sample t-test.
Abbreviations: DEXA, Dual-Energy X-ray Absorptiometry; SD, standard deviation. Bolded values are the statistically significant values.
Table 2.
Descriptive statistics of outcomes by individual grouping.
| Grouping by metabolic defect | |||||
|---|---|---|---|---|---|
|
Group A
(N = 268) |
Group B
(N = 11) |
Group C
(N = 1) |
Group D
(N = 10) |
Group E
(N = 4) |
|
| Sex, n (%) | |||||
| Female | 131 (48.9%) | 6 (54.5%) | 1 (100.0%) | 6 (60.0%) | 3 (75.0%) |
| Male | 137 (51.1%) | 5 (45.5%) | 0 (0.0%) | 4 (40.0%) | 1 (25.0%) |
| Race, n (%) | |||||
| White | 209 (78.0%) | 4 (36.4%) | 0 (0.0%) | 7 (70.0%) | 2 (50.0%) |
| Non-White | 59 (22.0%) | 7 (63.6) | 1 (100.%) | 2 (20.0%) | 13 (50.0%) |
| Autism, n (%) | 4 (1.5%) | 1 (9.1%) | 0 (0.0%) | 0 (0.0%) | 1 (25.0%) |
| Expressive language disorder/delay, n (%) | 1 (0.4%) | 3 (27.3%) | 0 (0.0%) | 1 (10.0%) | 0 (0.0%) |
| Blue sclera, n (%) | 221 (82.5%) | 7 (63.6%) | 1 (100.0%) | 5 (50.0%) | 2 (50.0%) |
| Dentinogenesis imperfecta, n (%) | 131 (48.9%) | 2 (18.2%) | 0 (0.0%) | 2 (20.0%) | 0 (0.0%) |
| Age at first hearing loss diagnosis, mean (SD) | 8.6 (6.16) | 2.0 (1.41) | 3.0 (..) | 8.0 (.) | |
| Age at first bisphosphonate infusion, mean (SD) | 3.9 (4.52) | 2.5 (4.61) | 3.0 (.) | 2.0 (1.76) | 3.8 (3.30) |
| Age at DEXA, mean (SD) | 11.4 (5.10) | 7.0 (4.84) | 8.0 (.) | 10.0 (4.16) | 11.5 (4.36) |
| Patient received bisphosphonates at time of DEXA, n (%) | 231 (87.5%) | 11 (100.0%) | 1 (100.0%) | 10 (100.0%) | 4 (100.0%) |
| DEXA Z-score, mean (SD) | −1.1 (1.73) | −1.7 (2.21) | −3.0 (.) | 0.4 (1.63) | −1.8 (1.41) |
| Family history of OI, n (%) | 91 (34.0%) | 0 (0.0%) | 0 (0.0%) | 3 (30.0%) | 0 (0.0%) |
| History of scoliosis, n (%) | 108 (40.4%) | 7 (63.6%) | 1 (100.0%) | 7 (70.0%) | 1 (25.0%) |
| Scoliosis operations, n (%) | 21 (19.4%) | 0 (0.0%) | 0 (0.0%) | 1 (14.3%) | 0 (0.0%) |
| History of rod placement for long bone deformity, n (%) | 184 (68.7%) | 9 (81.8%) | 1 (100.0%) | 7 (70.0%) | 3 (75.0%) |
| Presence of joint hypermobility, n (%) | 207 (77.2%) | 9 (81.8%) | 0 (0.0%) | 8 (80.0%) | 4 (100.0%) |
| Patient diagnosed short in stature, n (%) | 145 (54.1%) | 8 (72.7%) | 0 (0.0%) | 7 (70.0%) | 4 (100.0%) |
Descriptive statistics only for the comparison of Groups A-E due to small sample sizes.
Abbreviations: DEXA, Dual-Energy X-ray Absorptiometry; SD, standard deviation. The values were intentionally left blank as the N was too small for a SD.
OI genotype–phenotype relationships
Blue sclerae were significantly associated with genotype, being present in 83% and 58% of those with COL1A1/1A2 and non-COL1A1/A2 variants, respectively (p = 0.002). DI was also more common in the COL1A1/1A2 vs the non-COL1A1/A2 group (49% vs 15%, p < 0.001).
Similar proportions of enrolled patients in both the COL1A1/1A2 and the non-COL1A1/1A2 experienced hearing loss (15.7% vs 15.4%, respectively, p = 1.00). Participants in the COL1A1/1A2 group were more likely to be diagnosed with hearing loss later into childhood than participants in the non-COL1A1/1A2 group, but the difference was not significant (p = 0.13). Hearing loss was first diagnosed in the COL1A1/1A2 group at a median age of 7.0 (IQR: 3.0-14.0) while the non-COL1A1/1A2 group had a median diagnosis age of 3.0 (IQR: 2.0-5.5) yr old.
Most of the enrolled study participants have received bisphosphonate infusions (96.6%) and DEXA scans (99.0%). Both the COL1A1/1A2 and non-COL1A1/1A2 groups had high rates of bisphosphonate infusions (96.2% vs 100%, respectively). The median age at which the participant received their first bisphosphonate infusion did not significantly differ between the COL1A1/1A2 and non-COL1A1/1A2 groups (2.0 yr old [IQR: 0.0-6.0] vs 2.0 yr old [IQR: 0.0-3.0, p = 0.39]). The proportion of participants who had a DEXA scan was similar between the 2 groups, with 98.9% of COL1A1/1A2 and 100% of the non-COL1A1/1A2 participants. There was a significant difference in the average age of participants at their first DEXA scan between the COL1A1/1A2 and non-COL1A1/1A2 groups (11.4 yr [SD: 5.10] vs 8.9 yr [SD: 4.59, p = .02]). The results of the DEXA scans were not statistically significant (p = 0.68) between the 2 groups, with the mean Z-score for the COL1A1/1A2 group of –1.1 (SD: 1.73) and –0.9 (SD: 2.08) for the non-COL1A1/1A2 group. There was a moderate association in the proportion of patients receiving bisphosphonate infusions at the time of DEXA between the COL1A1/1A2 and non-COL1A1/1A2 groups (87.5% vs 100%, p = 0.05).
Nearly 70% of all patients had a history of rod placement for long bone deformity. Additionally, 42.3% of the participants in this study have scoliosis. Those in the non-COL1A1/1A2 group have significantly higher rates of scoliosis compared to those in the COL1A1/1A2 group (61.5% vs 40.4%, p = 0.04). However, it appears that the COL1A1/1A2 group has higher rates of scoliosis operations compared to the non-COL1A1/1A2 group, though not significant (19.4% vs 6.3%, p = 0.30).
The majority of all patients (55.8%) had a diagnosis of short stature; this was not significantly different between the groups (54.1% vs 73.1%, p = 0.06). Approximately 78% of all patients had joint hypermobility present, with no significant difference between the COL1A1/1A2 and non-COL1A1/1A2 groups (77.2% vs 80.8%, p = 0.68). Lastly, 34% of those in the COL1A1/1A2 group have a family history of OI, compared to only 12% of those in the non-COL1A1/1A2 group (p = 0.03).
Non-OI genotype phenotype relationships
Expressive language disorder/delay was significantly associated with COL1A1/1A2 grouping, with a prevalence of 15% and 0.4% in non-COL1A1/1A2 and COL1A1/1A2 patients, respectively (p < 0.001). ASD also had a higher, but insignificant, prevalence in non-COL groups compared to COL groups, though not significant (7.7% vs 1.5% respectively, p = 0.09). The presence of tics, seizures, and hydrocephalus is low across both groups (p = 1.00 for all 3 criteria). Tics were not reported in either grouping. Only one seizure was reported in the COL1A1/1A2 group. There were 3 reports of VPS in the COL1A1/1A2 group.
Discussion
The present study shows varying prevalence of certain phenotypic characteristics in OI when comparing individuals with COL1A1/1A2 to those with non-COL1A1/1A2 variants. To the best of our knowledge, this study utilizes one of the largest OI cohorts studied in the United States.
The presence or absence of blue sclera is a commonly studied finding of interest in those examining phenotypic differences in OI. We found that those with COL1A1/1A2 variants were significantly more likely (p = 0.002) to have blue sclera than those with non-COL1A1/1A2 variants. Similar findings have been reported in the prevalence of blue sclera but have often looked at different parameters. For instance, a study by Liu et al. in 2017 found that those with autosomal dominant inheritance, which are typically COL1A1/1A2 variants, were more likely to have blue sclera than those with autosomal recessive inheritance, typically of non-COL1A1/1A2 genes.13 Furthermore, several studies have shown that individuals with more mild clinical forms of the disease (OI type I), are significantly more likely to present with blue sclera than those with more severe clinical OI.13,15 While no clear etiology of this phenomenon has yet been described, we believe that this is likely due to the fact those with COL1A1/1A2 variants have structural collagen defects whereas those with non-COL1A1/1A2 variants usually have defects in the post-translational modification of bone and collagen.
We also found that individuals with COL1A1/1A2 variants were significantly more likely (p = 0.0009) to have co-existing DI compared to those with non-COL1A1/1A2 variants. While many OI genotype–phenotype studies have examined the relationship to DI, a clear relationship has not been defined. Some studies have shown that individuals with qualitative collagen defects are more likely to have DI,13 and others have found no difference in those with autosomal dominant inheritance when compared to those with autosomal recessive inheritance.13,16
Individuals with COL1A1/1A2 variants were significantly more likely (p = 0.03) to have a documented family history of OI compared to those with non-COL1A1/1A2 variants. This is consistent with the current understanding that those with COL1A1/1A2 variants are inherited in an autosomal dominant nature, and those with non-COL1A1/1A2 are inherited in an autosomal recessive nature.19
The significant difference in blue sclera, DI, family history of OI between patients with COL1A1/1A2 and non-COL1A1/1A2 variants can all inform the diagnosis, treatment, and management of OI patients based on their genetic variants. Blue sclera, DI, and family history are significantly more associated with having a COL1A1/1A2 variant than a non-COL1A1/1A2 variant. DI being more significant in COL1A1/1A2 shows a potential greater need for tailored dental care for COL1A1/1A2 patients. Blue sclera and DI can be visually perceptible hallmark signs of OI to primary care providers trying to distinguish between possible diagnoses and who may not encounter many OI patients regularly in their practice but are not definitive for making an OI diagnosis.
While this may suggest that there would be delayed treatment of patients with non-COL1A1/1A2 variants due to a significant difference in some of the more unique visual signs commonly attributed to OI, there was no significant difference in if patients had ever received a DEXA scan, the age of the patient during their first DEXA scan, or the Z-score from their DEXA scans suggesting no substantial delay in care is evident. Similarly, there was no significant difference in whether bisphosphonate therapy was used or age at first infusion.
There was a significant difference between the COL1A1/1A2 and non-COL1A1/1A2 groups regarding history of scoliosis, but no significant difference between the 2 groups on if they had received scoliosis surgery. This finding might suggest a benefit in more frequent or proactive screening for scoliosis for patients with non-COL1A1/1A2 variants for tracking scoliosis development, given that there is currently no standard of care regarding X-ray surveillance for scoliosis in patients with OI. Along the same line, there is no statistically significant difference between COL1A1/1A2 and non-COL1A1/1A2 groups and history of rod placement for long bone deformities. This would suggest that both groups should have the same screening and physical examination standards for assessing candidacy for rod placement surgery.
There was no significant difference between groups regarding the presence of hearing loss or the age at which they were first diagnosed with hearing loss, suggesting that similar screening practices should be maintained for both COL1A1/1A2 and non-COL1A1/1A2 patients. There was a significant difference however, in the development of expressive language disorder/delay with higher prevalence in the non-COL1A1/1A2 group. During annual OI standard of care visits, any suspected expressive language delay or disorders should be referred to a speech language pathologist for early intervention, with no considerations for variant type. Making providers aware that expressive language disorder/delay is more prevalent in patients with non-COL1A1/1A2 variants can be used to raise awareness and ensure timely intervention. There is no significant difference between COL1A1/1A2 groups and non-COL1A1/1A2 groups on the diagnosis of ASD. Similarly, any suspected ASD should be referred to the appropriate care provider for diagnosis and management.
There were few recorded instances of OI patients experiencing seizures, tics, or VPS in either group. It was initially thought that non-COL1A1/1A2 patients might have more neurological manifestations than COL1A1/1A2 patients since different genes are implicated in pathogenesis,20–23 but no difference was noted in the present analysis.
A patient's joint hypermobility and short stature can impact PT or OT recommendations or accommodations. The majority of both COL1A1/1A2 and non-COL1A1/1A2 patients have joint hypermobility, with no statistical difference between the 2 groups in prevalence. Similarly, there was no difference between the 2 groups in regard to being diagnosed with short stature. There being no difference between the 2 groups for both joint hypermobility and short stature would suggest no need to develop specified screening or management plans for patients belonging to either the COL1A1/1A2 or non-COL1A1/1A2 groups. Growth curves for COL1A1/1A2-related type III and IV OI already exist24; further evaluation of application of these curves for individuals with other genetic causes of severe and progressive OI are indicated.
While this cross-sectional study was still limited in its predictive power, comparing autosomal dominant and autosomal recessive forms of the disease due to a low number of patients in the non-COL1A1/1A2 group, it supported the prevalence of underlying genetic causes of OI with roughly 90% (n = 268) of our patients carrying COL1A1/1A2 variants which is consistent with well-established incidence data.1 As with other studies, the relatively low number of individuals with non-COL1A1/1A2 related OI limited the predictive power of this study. Furthermore, the lack of sex differences between the COL1A1/1A2 and non-COL1A1/1A2 suggests an even distribution of variants among males and females. In conclusion, this genotype/phenotype analysis in pediatric patients with OI shows similar distribution between the COL and non-COL patients. The increased incidence of expressive language disorder/delay is a particularly interesting finding and may warrant further exploration of concomitant genetic disorders or nearby genes that could contribute to this finding.
Acknowledgments
The authors are sincerely grateful to all the participants in this study.
Contributor Information
Jay J Byrd, College of Medicine, University of Nebraska Medical Center, Omaha, NE, 68198, United States; The Child Health Research Institute, University of Nebraska Medical Center, Omaha, NE, 68198, United States.
Andrew C White, College of Medicine, University of Nebraska Medical Center, Omaha, NE, 68198, United States.
Claire G Nissen, College of Medicine, University of Nebraska Medical Center, Omaha, NE, 68198, United States.
Makayla Schissel, Department of Biostatistics, College of Public Health, University of Nebraska Medical Center, Omaha, NE, 68198, United States.
Matthew Van Ormer, The Child Health Research Institute, University of Nebraska Medical Center, Omaha, NE, 68198, United States.
Danita Velasco, Children’s Nebraska, Department of Pediatrics, Division of Genetics Omaha, NE, 68114, United States.
Maegen Wallace, Phoenix Childrens Hospital, Department of Pediatric Orthopaedic Surgery Phoenix, AZ, 85016, United States.
Author contributions
Jay J. Byrd (Conceptualization, Methodology, Data acquisition, Formal analysis, Investigation, Writing—original draft, Writing—review & editing), Andrew C. White (Conceptualization, Methodology, Data acquisition, Formal analysis, Investigation, Writing—original draft, Writing—review & editing), Claire G. Nissen (Writing—original draft, Writing—review & editing), Makayla Schissel (Methodology, Formal analysis, Writing—original draft, Writing—review & editing), Matthew Van Ormer (Methodology, Project administration, Writing—review & editing), Danita Velasco (Conceptualization, Methodology, Investigation, Supervision, Writing—review & editing), and Maegen Wallace (Conceptualization, Methodology, Investigation, Supervision, Writing—review & editing).
Funding
Funding for this work was provided by the Child Health Research Institute at Children’s Nebraska and the University of Nebraska Medical Center.
Conflicts of interest
M.W. serves on the Osteogenesis Imperfecta Medical Advisory Council, is a board member for the Jansen’s Foundation, served on an advisory board for Ultragenyx. None of the above conflicted with this work.
Data availability
The data underlying this article cannot be shared publicly due to the sensitive nature of the data. The data will be shared on reasonable written request to the corresponding author.
References
- 1. Marini JC, Forlino A, Bächinger HP, et al. Osteogenesis imperfecta. Nat Rev Dis Prim. 2017;3(1):1-19. 10.1038/nrdp.2017.52 [DOI] [PubMed] [Google Scholar]
- 2. Choi Y, Hwang S, Kim GH, Lee BH, Yoo HW, Choi JH. Genotype-phenotype correlations and long-term efficacy of pamidronate therapy in patients with osteogenesis imperfecta. Ann Pediatr Endocrinol Metab. 2022;27(1):22-29. 10.6065/apem.2142144.072 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Barnes AM, Chang W, Morello R, et al. Deficiency of cartilage-associated protein in recessive lethal osteogenesis imperfecta. N Engl J Med. 2006;355(26):2757-2764. 10.1056/NEJMoa063804 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Franzone JM, Shah SA, Wallace MJ, Kruse RW. Osteogenesis imperfecta. Orthop Clin N Am. 2019;50(2):193-209. 10.1016/j.ocl.2018.10.003 [DOI] [PubMed] [Google Scholar]
- 5. Sillence DO, Senn A, Danks DM. Genetic heterogeneity in osteogenesis imperfecta. J Med Genet. 1979;16(2):101-116. 10.1136/jmg.16.2.101 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Bonafe L, Cormier-Daire V, Hall C, et al. Nosology and classification of genetic skeletal disorders: 2015 revision. Am J Med Genet A. 2015;167(12):2869-2892. 10.1002/ajmg.a.37365 [DOI] [PubMed] [Google Scholar]
- 7. Kang H, Aryal ACS, Marini JC. Osteogenesis imperfecta: new genes reveal novel mechanisms in bone dysplasia. Transl Res. 2017;181:27-48. 10.1016/j.trsl.2016.11.005 [DOI] [PubMed] [Google Scholar]
- 8. Unger S, Ferreira CR, Mortier GR, et al. Nosology of genetic skeletal disorders: 2023 revision. Am J Med Genet. 2023;191(5):1164-1209. 10.1002/ajmg.a.63132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Rauch F, Lalic L, Roughley P, Glorieux FH. Genotype–phenotype correlations in nonlethal osteogenesis imperfecta caused by mutations in the helical domain of collagen type I. Eur J Hum Genet. 2010;18(6):642-647. 10.1038/ejhg.2009.242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Rauch F, Lalic L, Roughley P, Glorieux FH. Relationship between genotype and skeletal phenotype in children and adolescents with osteogenesis imperfecta. J Bone Miner Res. 2010;25(6):1367-1374. 10.1359/jbmr.091109 [DOI] [PubMed] [Google Scholar]
- 11. Balasubramanian M, Parker MJ, Dalton A, et al. Genotype–phenotype study in type V osteogenesis imperfecta. Clin Dysmorphol. 2013;22(3):93-101. 10.1097/MCD.0b013e32836032f0 [DOI] [PubMed] [Google Scholar]
- 12. Lindahl K, Åström E, Rubin CJ, et al. Genetic epidemiology, prevalence, and genotype–phenotype correlations in the Swedish population with osteogenesis imperfecta. Eur J Hum Genet. 2015;23(8):1042-1050. 10.1038/ejhg.2015.81 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Liu Y, Asan MD, Ma D, et al. Gene mutation spectrum and genotype-phenotype correlation in a cohort of Chinese osteogenesis imperfecta patients revealed by targeted next generation sequencing. Osteoporos Int. 2017;28(10):2985-2995. 10.1007/s00198-017-4143-8 [DOI] [PubMed] [Google Scholar]
- 14. Mrosk J, Bhavani GS, Shah H, et al. Diagnostic strategies and genotype-phenotype correlation in a large Indian cohort of osteogenesis imperfecta. Bone. 2018;110:368-377. 10.1016/j.bone.2018.02.029 [DOI] [PubMed] [Google Scholar]
- 15. Maioli M, Gnoli M, Boarini M, et al. Genotype–phenotype correlation study in 364 osteogenesis imperfecta Italian patients. Eur J Hum Genet. 2019;27(7):1090-1100. 10.1038/s41431-019-0373-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Higuchi Y, Hasegawa K, Futagawa N, Yamashita M, Tanaka H, Tsukahara H. Genetic analysis in Japanese patients with osteogenesis imperfecta: genotype and phenotype spectra in 96 probands. Mol Genet Genomic Med. 2021;9(6):e1675-e1617. 10.1002/mgg3.1675 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Wei S, Yao Y, Shu M, et al. Genotype-phenotype relationship and follow-up analysis of a Chinese cohort with osteogenesis imperfecta. Endocr Pract. 2022;28(8):760-766. 10.1016/j.eprac.2022.05.003 [DOI] [PubMed] [Google Scholar]
- 18. Zerfu T, Yong B, Harrington J, Howard A. Does the skeletal phenotype of osteogenesis imperfecta differ for patients with non-COL1A1/2 mutations? A retrospective study in 113 patients. J Pediatr Orthop. 2022;42(5):e507-e514. 10.1097/BPO.0000000000002116 [DOI] [PubMed] [Google Scholar]
- 19. Forlino A, Marini JC. Osteogenesis imperfecta. Lancet. 2016;387(10028):1657-1671. 10.1016/S0140-6736(15)00728-X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Fahiminiya S, Majewski J, Mort J, Moffatt P, Glorieux FH, Rauch F. Mutations in WNT1 are a cause of osteogenesis imperfecta. J Med Genet. 2013;50(5):345-348. 10.1136/jmedgenet-2013-101567 [DOI] [PubMed] [Google Scholar]
- 21. Keupp K, Beleggia F, Kayserili H, et al. Mutations in WNT1 cause different forms of bone fragility. Am J Hum Genet. 2013;92(4):565-574. 10.1016/j.ajhg.2013.02.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Pyott SM, Tran TT, Leistritz DF, et al. WNT1 mutations in families affected by moderately severe and progressive recessive osteogenesis imperfecta. Am J Hum Genet. 2013;92(4):590-597. 10.1016/j.ajhg.2013.02.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Aldinger KA, Mendelsohn NJ, Chung BH, et al. Variable brain phenotype primarily affects the brainstem and cerebellum in patients with osteogenesis imperfecta caused by recessive WNT1 mutations. J Med Genet. 2016;53(6):427-430. 10.1136/jmedgenet-2015-103476 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Barber LA, Abbott C, Nakhate V, Do AND, Blissett AR, Marini JC. Longitudinal growth curves for children with classical osteogenesis imperfecta (types III and IV) caused by structural pathogenic variants in type I collagen. Genet Med. 2019;21(5):1233-1239. 10.1038/s41436-018-0307-y [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.
Data Availability Statement
The data underlying this article cannot be shared publicly due to the sensitive nature of the data. The data will be shared on reasonable written request to the corresponding author.