Alendronate

Clinical, genetic characteristics and treatment outcomes of children and adolescents with osteogenesis imperfecta: a two-center experience

İbrahim Mert Erbaş a, Deniz İlgün Gürelb, Zehra Manav Kabayeğit c, Altuğ Koç d, Tolga Ünüvare,
Ayhan Abacı a, Ece Böber a, and Ahmet Anık e
aDivision of Pediatric Endocrinology, Faculty of Medicine, Dokuz Eylül University, İzmir, Turkey; bDepartment of Pediatrics, Faculty of Medicine, Aydın Adnan Menderes University, Aydın, Turkey; cDepartment of Medical Genetics, Faculty of Medicine, Aydın Adnan Menderes University, Aydın, Turkey; dDepartment of Medical Genetics, Faculty of Medicine, Dokuz Eylül University, İzmir, Turkey; eDivision of Pediatric Endocrinology, Faculty of Medicine, Aydın Adnan Menderes University, Aydın, Turkey

ARTICLE HISTORY
Received 20 November 2020
Accepted 17 May 2021 Published online 05 June 2021

ABSTRACT

Background: Osteogenesis imperfecta (OI), is a heritable, heterogeneous connective tissue dis- order, characterized by fragile bones. There are conflicting results about genotype-phenotype correlations and efficiency of bisphosphonate treatment in this disorder.
Aim: We aimed to evaluate the clinical, genetic characteristics, and long-term follow-up results of children and adolescents with OI.
Materials and methods: A two-center retrospective study was conducted using demographic, clinical, and genetic data obtained from the medical records of the patients.
Results: Twenty-nine patients (62% male, median age; 3.6 years) with OI diagnosis from 26 families were included in the study. Thirteen different variants (nine were novel) were described in 16 patients in COL1A1, COL1A2, and P3H1 genes. Our siblings with homozygous P3H1 variants had a severe phenotype with intrauterine and neonatal fractures. Twenty-two patients were treated with bisphosphonates (17 of them with pamidronate, five with alendronate) with a median duration of 3.0 (1.6–4.8) years. Eleven patients (50%) suffered from fractures after the treatment. Haploinsufficiency variants in COL1A1 caused a milder skeletal phenotype with less fracture count and better treatment outcomes than structural variants. When compared with the anthropometric measurements at the initial diagnosis time, height Z-scores were lower on the last clinical follow-up (p = 0.009).
Conclusions: We could not find an obvious genotype-phenotype correlation in Turkish OI patients with COL1A1 or COL1A2 variants. Treatment with pamidronate was effective in reducing
fracture counts, without any long-term adverse effects.

KEYWORDS
Anti-resorptive drugs; collagen; P3H1; osteogenesis imperfecta; pediatrics

Introduction

Osteogenesis imperfecta (OI), also known as brittle bone disease, is a heritable, heterogeneous connective tissue disorder, characterized by fragile bones1. Also, other organs and systems can be involved due to the disorder of the extracellular matrix. Dentinogenesis imperfecta, soft tissue abnormalities, such as discolora- tion of the sclera, or hearing loss are among the fre- quent clinical features2. OI affects approximately 1 in 10,000 to 20,000 live births and the incidence varies according to the region3.
The primary cause of OI is variants in COL1A1 (OMIM 120,150) or COL1A2 (OMIM 120,160), which encodes α1 and α2 chains of type I collagen1,3. Nearly 2000 dominant variants have been reported in these genes, causing a variety of skeletal phenotypes with different severity. More than 85% of variants are heterozygous and can be inherited from an affected parent in an autosomal dominant manner or arise de novo4,5. Recently, recessive OI forms have also been identified, caused by variants in several genes including P3H1 (OMIM: 610,915; formerly LEPRE1), whose pro- ducts influence the biosynthesis, structure, or function of type I collagen. New gene discoveries in this subject have provided a better explanation for the pathophy- siology of OI6.
Treatment of patients with OI should be individua- lized consistent with age, the severity of the disease, or functional conditions. The objectives of treatment are to improve quality of life and to decrease fracture counts7. Bone anti-resorptive drugs are the most com- monly recommended treatment approaches for OI, but there is a lack of consensus on treatment strategies. Intravenous bisphosphonate treatment has been the most widely preferred way to ameliorate bone fragility in children and adolescents for the last 30 years. However, the efficiency of this treatment method and long-term follow-up results remains controversial7,8.
In this study, we aimed to evaluate the clinical, genetic characteristics and long-term follow-up results of children and adolescents with OI, and the efficacy of bisphosphonates according to underlying genotype. To the best of our knowledge, there is no study focusing on the genotype-phenotype correlation in the Turkish pediatric population with OI. In addition, findings from this study are expected to facilitate better medical care and genetic counseling by enriching the genotype and phenotype databases in such rare clinical conditions.

Materials and methods

A two-center retrospective study was conducted, including patients aged 0–18 years with OI, who were referred to participating pediatric endocrinology out- patient clinics between 2008 and 2020. OI diagnosis was defined according to a conventional classification system, based on the clinical findings9. Firstly, a total of 29 OI patients with available data were identified by clinical features. Second, a search for variants of COL1A1 and COL1A2, as reference genes for OI, led to the identification of 14 patients with variants. Third, the P3H1 gene was searched as a further test in two siblings, who had consanguinity between the parents and did not show a variant in COL1A1 or COL1A2. In the remaining 13 patients, a variant could not be observed in COL1A1 or COL1A2 (n = 6), or genetic analysis could not be performed due to the lack of parental consent, or some of them did not come to follow-up.
Data including the birth history, age at diagnosis, clinical presentations, fracture histories (time of the first fracture, total number, and location of fractures), treatment modalities, and durations were collected from the medical records of the patients, which were filled during routine clinical visits. All fractures were confirmed by radiographs. All sclera colors on the blue-gray scale were described as blue. Height was measured with the Harpenden stadiometer with an accuracy of 0.1 cm for those who could stand, or the crown–heel length was used for those who could not stand. Bodyweight was measured with a SECA scale with an accuracy of 0.1 kg, with the patient in light clothing. Z-scores for weight and height were calcu- lated with the online calculator for pediatric endocri- nologists (Child Metrics)10, using the references created for the Turkish population by Neyzi et al.11.
Serum electrolytes, parathyroid hormone, alkaline phosphatase, and 25-OH vitamin D levels on the first admission were noted. Clinical and biochemical changes were evaluated regularly every 3 months dur- ing the follow-up.
Clinical phenotypes of OI were defined according to the classification of the International Nomenclature Committee for Constitutional Disorders of the Skeleton12. Patients with type I or IV OI, without intrauterine fractures and who had less than or equal to one fracture per year and fully ambulant other than at times of acute fracture were admitted as mild phe- notype. Patients with type III OI, who showed marked impairment of linear growth, or who had a progressive deformity of long bones or spines and wheel-chair dependent, those with congenital fractures or more than three fractures per year were defined as severe phenotype12.
Lumbar (L1–L4) bone mineral density (BMD) was evaluated by dual-energy X-ray absorptiometry (DXA) in patients older than one-year-old, with Z-score values adjusted for age and gender for the Turkish population13. A Z-score of less than or equal to −2.0 accepted as low. BMD after the treatment was evaluated at least 1 year after the treatment.
Treatment with pamidronate was performed by infusions over 3–4 hours for every three months, dos- ing according to age14. A treatment cycle delineated a 2-day administration period. Oral treatment with alendronate was given with a weekly dose of 35 mg in patients weighing ≤30 kg, and 70 mg in those >30 kg. A clinical decision based on the severity of phenotype and number of fractures was used for initiating the treatment. Patients with type I or IV OI, who had a history of two or more long-bone fractures in the last year, and all patients with type III OI were given bisphosphonate treatment. Patients with a BMD in the normal range or with an exceedingly mild phenotype or without any fractures in the last year before admission were not treated15. Treatment responsiveness was defined according to changes in the Z-score of BMD or fractures that occurred after the initiation of treatment16.
Variant analysis of the COL1A1, COL1A2, and P3H1 genes was performed with Sanger sequencing, using a sample obtained from peripheral blood. All exons and flanking sequences were amplified by polymerase- chain reaction (PCR) using previously described pri- mers. PCR products were analyzed after the thermal cycling program and purification steps.
This study was approved by the ethical committee of Aydın Adnan Menderes University Faculty of Medicine (Approval number: 2020/226-7) and an informed written consent form was not obtained due to the retrospective nature of the study.

Statistical analyses
All statistical analyses were performed using the SPSS application for Windows version 24.0 (IBM Corp. Released 2016. Armonk, NY: IBM Corp.). Clinical data were presented as number (%) for categorical and median (with the respective 25–75th percentile values) for numeric values. The Wilcoxon signed-rank test was used to analyze continuous variables among dependent groups. A p-value of <0.05 was considered statistically significant. Results Twenty-nine patients (62% male) with OI diagnosis from 26 families were included in the study, whose median birth age was 39.0 (38.0–39.5) weeks. Fifteen, eight, and six patients were classified as OI type I, III, and IV, respectively. The median age of diagnosis was 3.6 (1.0–9.8) years. Median weight and height of the patients were 14 (5.9–28.5) kg (Z-score; −1.3 [(−2.0)— (−0.8)]) and 98 (62–135) cm (Z-score; −1.4 [(−2.2)— (−0.5)]), respectively. In the physical examination, blue sclera was noted in 79% of the patients (88% in those with a variant). Dentinogenesis imperfecta was observed in 24% of the patients (38% in those with a variant) on the follow-up. None of the patients had hearing loss. Demographic and clinical characteristics of the patients with a variant were shown in Tables 1 and 2. The median age of the first fracture was 1.5 (1.0–3.0) years. Four patients had intrauterine fractures. All patients had a median rate of 0.7 (0.4–2.0) fractures per year, whereas this was found as 1.0 (0.5–2.0) frac- tures per year in patients with a variant (Table 1). The most common affected bone before treatment was tibia in 52% of patients, followed by femur (41%). One patient with a variant (Case 5) had no fractures and was diagnosed by a genetic screening test because of her sibling (Case 6). Although they both presented the same variant, clinical presentations of the disease were different. Patients were followed up for a median period of 3.6 (1.5–5.2) years. On the last follow-up, the median age of the patients with available data was 7.4 (5.1–12.9) years. Weight and height Z-scores were −1.4 [(−3.4)— (−0.6)] and −2.2 [(−3.5)—(−0.8)], respectively. When compared with the anthropometric measurements at the time of initial diagnosis, weight Z-scores were simi- lar on the last clinical follow-up (p = 0.449). Fourteen patients (48%) had short stature on the last examina- tion and height Z-scores were significantly lower than the initial evaluation (p = 0.009). Bone mineral density Z-score was < −2 before the treatment in 47% of the patients with available data (n = 17), with a median Z-score of −1.6 [(−3.0)— (−0.6)] in all of them and −2.7 [(−3.2)—(−0.8)] in those with a variant. Laboratory results were in normal ranges with median levels of serum calcium and phos- phorus 9.8 (9.2–10.2) and 4.8 (4.4–5.2) mg/dL, respec- tively. Serum alkaline phosphatase, parathyroid hormone, and 25-OH vitamin D levels were 236.5 (206.5–292.5) IU/L, 25.9 (14.0–42.0) pg/mL, and 18.3 (13.5–35.2) ng/mL, respectively. None of the patients had nephrocalcinosis. In genetic analyses, 13 different variants were described in 16 patients in COL1A1, COL1A2, and P3H1 genes, while 69% (n = 9) of them were novel. The most common affected amino acid was glycine in 31% (n = 4) of the variants. Thirteen patients (81%) had a heterozygous variant in COL1A1. The frequency of missense, nonsense, and frameshift variants were 64%, 18%, and 18%, respectively. One patient had a heterozygous missense variant in COL1A2 and two siblings had a homozygous missense variant in P3H1 (Table 1). Twenty-two patients (13 with a variant) were treated with bisphosphonates with a median duration of 3.0 (1.6–4.8) years. Seventeen of them (77%) were treated with pamidronate, and five (23%) with alendronate. Four of the pamidronate-treated patients (24%) had a fever on the first administration. Seven of the pami- dronate-treated (41%) and four of the alendronate- treated (80%) patients suffered from fractures after the treatment. A total of 0.2 (0.0–0.7) fractures per year was observed in treated patients [0.0 (0.0–0.6) fractures per year in individuals with a variant] during a median follow-up period for 3.8 (2.6–6.5) years. All alendronate-treated patients suf- fered a fracture within the first six months after the initiation of treatment. No long-term adverse effects were observed with bisphosphonates in our OI patients. In those with an underlying genetic cause, six patients suffered from fractures after the treatment; one of them had a variant in COL1A2 (Case 7), one of them in P3H1 (Case 15, 16), and the remaining ones in COL1A1 (Case 10, 13, 14). Two of the variants included a glycine substitution in COL1A1 (Case 7, 14). Case 13, 14, and 15 had intrauterine fractures and started treatment in the infantile period. Case 10 was treated with alendronate and suffered an early fracture two months after the initiation of treatment. Table 1. Demographic, clinical, and genetic characteristics of the patients with osteogenesis imperfecta, who showed a variant in searched genes. Family Case Gender Birth age (week) Consanguineous marriage Family history of OI Blue sclera DI Additional clinical features Age of the first fracture Numbers* and locations of fractures before the presentation Type of OI Results of the genetic analysis Gene Variant Type of the Expression of the The novelty DI: dentinogenesis imperfecta, OI: osteogenesis imperfecta, F: female, M: male, IU: Intrauterine, Cl: Clavicula, Co: Costa, Fe: Femur, H: Humerus, P: Phalanx, R: Radius, T: Tibia, U: Ulna, V: Vertebrae, N/A: not available. *Numbers of the fractures were presented as fractures per year.mTreatment characteristics and outcomes of the patients with a variant were presented in Table 2. Discussion In this study, we presented 29 patients with OI diag- nosis from 26 families, and 13 different variants (nine were novel) were described in 16 patients in COL1A1, COL1A2, and P3H1 genes. OI is one of the most com- mon skeletal dysplasias, caused by a deficiency or defective production in type I collagen chains α1 and α2. Variants in either COL1A1 or COL1A2 correspond to 40–90% of patients with OI, depending on the population. Also, COL1A1 variants were more fre- quently reported than COL1A217. The remaining patients have shown various genetic causes, but similar phenotypic features with individuals who had patho- genic variants in COL1A1 or COL1A218. OI is usually inherited in an autosomal dominant manner, but auto- somal recessive patterns have also been reported12. Missense variants in COL1A1 or COL1A2 were found as the most common type of variants in patients with OI19. The prevalence of familial patients ranged between 32–53% in the literature, which varied accord- ing to countries20–22. Lin et al.19 reported the rate of novel variants as 41% in their study. In this study, 23% of the variants were familial and we documented nine (69%) novel variants in patients with OI. All variants in COL1A1 and COL1A2 were heterozygous, pointing to the dominant inheritance pattern of this disease. The homozygous variant was detected only in P3H1. This finding supports that the autosomal recessive pattern of this clinical condition is related to other genes rather than COL1A1 or COL1A2. Our results were familiar with the literature data and the variants were family- specific, which showed that our patient group had no hotspot variant in the described genes. Also, novel pathogenic collagen type I variants enrich the genetic database of OI and contributes to the understanding of genotype-phenotype correlations for this disorder. Genotype-phenotype correlations in OI have been investigated for many years and some studies have shown a significant correlation23,24. More severe phe- notypes were reported in patients with COL1A1 patho- genic variants, than COL1A225. Mrosk et al.26 claimed that there was a strong correlation between genotype and severe phenotypes, and they made a ranking according to phenotype severity as follows: P3H1, COL1A1, and COL1A2, respectively. On the other hand, the genotype-phenotype relationship remains unclear, while various carriers of the same variant might develop diverse phenotypes and the circum- stances affecting additional phenotype modification have not been elucidated yet27. This may be related to the incomplete penetrance characteristic of this disorder12. Severe OI may present with in utero frac- tures and can be detected on prenatal ultrasound examination18. Moderate forms of the disease can cause progressive bone deformities, such as scoliosis, rib cage deformities, or bowing of long bones. Short stature and reduced growth velocity can also be seen in milder forms28. In this study, five (31%) of the patients with a variant had a severe phenotype and were diag- nosed with fractures that occurred in the intrauterine or neonatal period. These severe phenotypes were asso- ciated with variants in COL1A1, COL1A2, and P3H1. However, most of the patients with COL1A1 variants had a mild to moderate phenotype. Our findings sug- gest that P3H1variants cause a severe phenotype in OI as reported before. But, genotype-phenotype correla- tions did not only depend on whether the causative gene is COL1A1 or COL1A2. We presented a patient without a history of fractures, who was diagnosed by the family screening and had the same variant as her sibling. This may be explained by not only the genetic structure, but also a multifactorial process that may play a significant role in the pathophysiology of OI. Nevertheless, such a limited sample size in this study might be insufficient to find an exact genotype- phenotype correlation. In previous studies, lower limbs were reported as the most affected body compartment in OI patients. This result was interpreted to be related to the greater load on the lower extremities25,29. Similar to this finding, we observed the highest number of fractures in the lower limbs. The frequency of dentinogenesis imperfecta and blue sclera in OI cohorts ranged between 25–61% and 80–93%, respectively19,25,29,30. Rauch et al.31 demon- strated that all OI patients with variants in the amino- terminal end of the α1 collagen chain had blue sclera. Nevertheless, in another study, the presence of blue sclera varied among patients with the same variant in COL1A1 or COL1A232. In our OI patients, 79% had blue sclera (88% in those with a variant), a finding independent of the location or type of the variant. There have been conflicting results about the correla- tion of genotype and dental phenotype, as well25. Glycine substitutions in α1 and α2 chains of collagen were associated with dentinogenesis imperfecta, while amino-terminal variants were not16,31. Rauch et al.31 showed that none of their OI patients were diagnosed with dentinogenesis imperfecta, who had glycine var- iants in the first 123 and 127 amino acids in COL1A1 and COL1A2, respectively. In contrast to previous stu- dies, we found dentinogenesis imperfecta in patients with variants as early as in the first 45 amino acids. Table 2. Anthropometric measurements, treatment characteristics, and outcomes of the patients with osteogenesis imperfecta, who showed a variant in searched genes. Family Case BMD before the treatment [g/ cm2 (Z-score)] Treatment agent Dosage of the treatment (mg/kg/ Adverse effects Treatment duration (year) Numbers* of the fractures after treatment Time of the first fracture after the treatment BMD after the treatment [g/ cm2 (Z-score)] Does the treatment continue? Anthropometric measurements on the first examination Age at Weight Height Anthropometric measurements on the last examination Age Weight Height BMD: Bone mineral density, N/A: not available. *Numbers of the fractures were presented as fractures per year during the follow-up period. Also, we could not find an exact relationship between dentinogenesis imperfecta and the type of affected amino acid, whether it was glycine or another substitu- tion. However, siblings sharing the same variant had a high concordance for the presence of dentinogenesis imperfecta or blue sclera, suggesting that the under- lying collagen variant is prominent for these clinical features in children with OI. There are two main categories of collagen variants resulting in dominant OI; haploinsufficiency and struc- tural. Frameshift, nonsense, and splice-site variants in COL1A1 or COL1A2 lead to a quantitative defect resulting in haploinsufficiency of collagen chain pro- duction. However, structural and qualitative defects are caused by missense variants, especially by the replace- ment of glycine by another amino acid in the Gly- X-Y triplet domain of the triple helix form. A milder phenotype and bone fragility were reported in indivi- duals with quantitative defects than those with qualita- tive variants in COL1A1 or COL1A219,20,23,25,27. Lin et al.19 found that OI patients with haploinsufficiency variants had more severe phenotypes, including lower weight, height, and BMD, when compared with quali- tative variants. However, patients with glycine struc- tural substitutions presented mild phenotypic features in several studies25,27. Serine and glycine substitutions were known as the most common type of missense variants in both COL1A1 and COL1A2, while we found glycine as the most affected amino acid in this study19,27. Although Kanno et al.16 reported that serine for glycine replacement was the severest variant type, we could not observe this finding in a patient with a similar substitution, who had normal linear growth and was responsive to the treatment. Our results were consistent with the literature data, showing that patients with haploinsufficiency variants in COL1A1 had a milder skeletal phenotype with a lower fracture count and better treatment outcomes than those with structural variants. In addition, many molecular and physical factors that alter bone morphology and strength may potentially affect the phenotypic reflec- tion of collagen gene variants, making a precise pheno- type-genotype correlation unattainable. P3H1 encodes prolyl 3-hydroxylase, an enzyme that plays a role in collagen modification. Variants in P3H1 cause a rare and severe to lethal OI phenotype, with excess hydroxylation and subsequent glycosylation of the helical region in collagen chains6,33,34. It is also expressed in the cartilage tissue, and thus, affected patients present osteochondrodystrophy with severe rhizomelia and neonatal fractures. Individuals with P3H1 variants had white sclera, growth deficiency, and very low BMD, in the previous studies33,35,36. Hearing loss and dentinogenesis imperfecta were absent in reported OI patients with P3H1 variants33,36. Our siblings with homozygous P3H1 variants had a severe phenotype with intrauterine and neonatal frac- tures, as reported in the literature. However, to the best of our knowledge, we described the first OI children with a P3H1 variant, who presented blue sclera and dentinogenesis imperfecta. This reflects the lack of knowledge about genotype-phenotype correlation in P3H1 variants, due to the low number of patients around the world. Therefore, this rare genetic cause of OI remains a focus for further researches. Bisphosphonates, which are anti-resorptive drugs, have become traditional treatment agents for OI. Pamidronate is the most frequently used bisphosphonate for this diagnosis and has a clear impact on increasing BMD, ameliorating linear growth, and decreasing fracture rates in affected individuals37–39. Dwan et al.40 summar- ized improvements of 1.5 Z-scores in BMD of children with OI in the first year of the bisphosphonate treatment. Some studies demonstrated the maximum benefit of bisphosphonates on BMD after 2–3 years of treatment38,41. There were suggestions to pause treatment at the end of 3 years for suitable patients, with a close follow-up for new fractures42. Bains et al.43 showed that bisphosphonate therapy was associated with lower frac- ture numbers. Shi et al.44 found a diminution in the risk of fractures in children with OI, but not in adult patients. It has also been reported that fracture counts reduced to 0.6 fractures per patient per year from 3.2 fractures, with a decrease of about 75% by pamidronate45. Nevertheless, there have been doubts about a reduction of long bone fractures by this treatment method, as shown in meta- analyses40,46,47. This may be related to the non-dynamic bone structure formed with the inhibition of osteoclasts by bisphosphonates, causing unrepaired microcracks48. Also, a reduction in bone quality was observed with bisphosphonate administration in animal studies49. Despite the conflicting literature data, we found that 59% of children treated with pamidronate had no further fractures after the treatment was initiated, on the follow- up for a median of 3.8 years. The median fracture rate per year per patient was decreased from 0.7 to 0.2 with bisphosphonate treatment but the length of the follow-up period was not standard for all patients. Although treat- ment should be individualized for children with OI, we observed that one or even 4 years of treatment did not differ in the possibility of further fractures in mild phe- notypes. In addition, patients with the P3H1 variant in this study showed a low response to the treatment. The positive influence of intravenous pamidro- nate has not been observed with oral bisphospho- nates, such as alendronate50. While alendronate did not reduce fracture numbers in moderate or severe OI patients, its benefits were found to be limited to mild phenotypes only7,50. Similar to these findings, we used alendronate only in patients with a mild phenotype; however, 80% of them suffered a fracture within the first 6 months after the initia- tion of treatment. This may be related to the later- onset efficacy of oral bisphosphonates on bone structure than intravenous ones. However, since we did not have a relatively large group of patients treated with oral bisphosphonates, we could not clarify this thought. To our knowledge, this is the first study focusing on the genotype-phenotype correlation in Turkish children with OI. However, this study has some lim- itations. First of all, it was a retrospective study and the biggest limitation was the sample size. Although the limited number of participants reflects the infre- quency of this disorder in the population, this made it difficult to draw any exact interpretation about genotype-phenotype correlations. Secondly, a DXA scan could not be performed on all patients after treatment. Thus, we used clinical outcomes such as the recurrence of fractures to evaluate the respon- siveness to treatment. However, this was considered equal to BMD as a measure of treatment outcome in previous long-term follow-up studies, whereas there is no useful marker to interpret bone quality. Finally, due to the retrospective nature of the study, we did not have the longitudinal data about teeth develop- ment from milk to permanent in the same indivi- duals, which may have affected our results regarding dentinogenesis imperfecta. Therefore, further, multi- center studies with larger cohorts and longer follow- up periods, including molecular researches for patho- physiologic pathways are warranted in this subject. More participants should be included in order to make it possible to reveal accurate genotype- phenotype correlations. This study demonstrated eight novel variants in COL1A1 and COL1A2 causing OI in children, and two siblings with a novel variant in P3H1. Although the P3H1 variant resulted in a severe phenotype, we could not find an obvious genotype-phenotype correla- tion in OI patients with either COL1A1 or COL1A2 variants. However, haploinsufficiency variants in COL1A1 caused a milder skeletal phenotype with a less fracture count and better treatment outcomes, than structural variants. Finally, 59% of the children treated with pamidronate had no further fractures after the initiation of treatment, without any long-term adverse effects. Acknowledgments None. Disclosure of potential conflicts of interest No potential conflict of interest was reported by the author(s). Funding The authors have no funding to report. ORCID İbrahim Mert Erbaş http://orcid.org/0000-0001-9368-8868 Zehra Manav Kabayeğit http://orcid.org/0000-0002-9505- 0371 Altuğ Koç http://orcid.org/0000-0002-8366-6806 Ayhan Abacı http://orcid.org/0000-0002-1812-0321 Ece Böber http://orcid.org/0000-0001-8828-0892 Ahmet Anık http://orcid.org/0000-0002-7729-7872 References 1. Rauch F, Glorieux FH.Osteogenesis imperfecta. Lancet. 2004;363(9418):1377–1385.doi:10.1016/S0140-6736(04) 16051-0. 2. Trejo P, Rauch F.Osteogenesis imperfecta in children and adolescents—new developments in diagnosis and treatment. Osteoporos Int. 2016;27(12):3427–3437. doi:10.1007/s00198-016-3723-3. 3. Lim J, Grafe I, Alexander S, Lee B. Genetic causes and mechanisms of osteogenesis imperfecta. Bone. 2017;102:40–49.doi:10.1016/j.bone.2017.02.004. 4. Lamandé SR, Bateman JF.Genetic disorders of the extracellular matrix. Anat Rec (Hoboken). 2020;303 (6):1527–1542.doi:10.1002/ar.24086. 5. Cohen JS. Patterns of inheritance in osteogenesis imperfecta. In: Shapiro JR, Byers PH, Glorieux FH, Sponseller PD, editors. Osteogenesis imperfecta: a translational approach to brittle bone disease. 1st ed. New York: Elsevier; 2013. p. 99. 6. Forlino A, Cabral WA, Barnes AM, Marini JC.New per- spectives on osteogenesis imperfecta. Nat Rev Endocrinol. 2011;7(9):540–557.doi:10.1038/nrendo.2011.81. 7. Palomo T, Vilaça T, Lazaretti-Castro M.Osteogenesis imperfecta: diagnosis and treatment. Curr Opin Endocrinol Diabetes Obes. 2017;24(6):381–388. doi:10.1097/MED.0000000000000367. 8. Biggin A, Munns CF.Long-term bisphosphonate ther- apy in osteogenesis imperfecta. Curr Osteoporos Rep. 2017;15(5):412–418.doi:10.1007/s11914-017-0401-0. 9. Sillence DO, Senn A, Danks DM.Genetic heterogeneity in osteogenesis imperfecta. J Med Genet. 1979;16 (2):101–116.doi:10.1136/jmg.16.2.101. 10. Demir K, Konakçı E, Özkaya G, Kasap Demir B, Özen S, Aydın M, Darendeliler F.New features for Child Metrics: further growth references and blood pressure calculations. J Clin Res Pediatr Endocrinol. 2020;12(2):125–129.doi:10.4274/jcrpe. galenos.2019.2019.0127. 11. Neyzi O, Bundak R, Gökçay G, Günöz H, Furman A, Darendeliler F, Baş F.Reference values for weight, height, head circumference, and body mass index in Turkish children. J Clin Res Pediatr Endocrinol. 2015;7 (4):280–293.doi:10.4274/jcrpe.2183. 12. Van Dijk FS, Sillence DO.Osteogenesis imperfecta: clinical diagnosis, nomenclature and severity assessment. Am J Med Genet A. 2014;164 (6):1470–1481.doi:10.1002/ajmg.a.36545. 13. Goksen D, Darcan S, Coker M, Kose T.Bone mineral density of healthy Turkish children and adolescents. J Clin Densitom. 2006;9(1):84–90.doi:10.1016/j. jocd.2005.08.001. 14. Glorieux FH.Osteogenesis imperfecta. Best Pract Res Clin Rheumatol. 2008;22(1):85–100.doi:10.1016/j. berh.2007.12.012. 15. Biggin A, Munns CF.Osteogenesis imperfecta: diagno- sis and treatment. Curr Osteoporos Rep. 2014;12 (3):279–288.doi:10.1007/s11914-014-0225-0. 16. Kanno J, Saito-Hakoda A, Kure S, Fujiwara I. Responsiveness to pamidronate treatment is not related to the genotype of type I collagen in patients with osteogenesis imperfecta. J Bone Miner Metab. 2018;36 (3):344–351.doi:10.1007/s00774-017-0840-9. 17. Kataoka K, Ogura E, Hasegawa K, Inoue M, Seino Y, Morishima T, Tanaka H.Mutations in type I collagen genes in Japanese osteogenesis imperfecta patients. Pediatr Int. 2007;49(5):564–569.doi:10.1111/j.1442- 200X.2007.02422.x. 18. Marini JC, Forlino A, Bächinger HP, Bishop NJ, Byers PH, Paepe A, Fassier F, Fratzl-Zelman N, Kozloff KM, Krakow D, Montpetit K, Semler O. Osteogenesis imperfecta. Nat Rev Dis Primers. 2017;3:17052. 19. Lin HY, Chuang CK, Su YN, Chen MR, Chiu HC, Niu DM, Lin SP.Genotype and phenotype analysis of Taiwanese patients with osteogenesis imperfecta. Orphanet J Rare Dis. 2015;10(1):152.doi:10.1186/ s13023-015-0370-2. 20. Zhang ZL, Zhang H, Ke YH, Yue H, Xiao WJ, Yu JB, Gu JM, Hu WW, Wang C, He JW, Fu WZ.The identi- fication of novel mutations in COL1A1, COL1A2, and LEPRE1 genes in Chinese patients with osteogenesis imperfecta. J Bone Miner Metab. 2012;30(1):69–77. doi:10.1007/s00774-011-0284-6. 21. Venturi G, Tedeschi E, Mottes M, Valli M, Camilot M, Viglio S, Antoniazzi F, Tatò L.Osteogenesis imperfecta: clinical, biochemical and molecular findings. Clin Genet.2006;70(2):131–139.doi:10.1111/j.1399- 0004.2006.00646.x. 22. Lee KS, Song HR, Cho TJ, Kim HJ, Lee T-M, Jin H-S, Park H-Y, Kang S, Jung S-C, Koo SK, et al. . Mutational spectrum of type I collagen genes in Korean patients with osteogenesis imperfecta. Hum Mutat. 2006;27(6):599.doi:10.1002/humu.9423. 23. 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. doi:10.1359/jbmr.091109. 24. Ben Amor IM, Glorieux FH, Rauch F. Genotype- phenotype correlations in autosomal dominant osteo- genesis imperfecta. J Osteoporos. 2011;2011:540178. doi:10.4061/2011/540178. 25. Zhytnik L, Maasalu K, Pashenko A, Khmyzov S, Reimann E, Prans E, Kõks S, Märtson A. COL1A1/2 pathogenic variants and phenotype characteristics in ukrainian osteogenesis imperfecta patients. Front Genet. 2019;10:722.doi:10.3389/fgene.2019.00722. 26. Mrosk J, Bhavani GS, Shah H, Hecht J, Krüger U, Shukla A, Kornak U, Girisha KM. Diagnostic strategies and genotype-phenotype correlation in a large Indian cohort of osteogenesis imperfecta. Bone. 2018;110:368–377.doi:10.1016/j.bone.2018.02.029. 27. Marini JC, Forlino A, Cabral WA, Barnes AM, San Antonio JD, Milgrom S, Hyland JC, Körkkö J, Prockop DJ, De Paepe A, Coucke P, Symoens S, Glorieux FH, Roughley PJ, Lund AM, Kuurila-Svahn K, Hartikka H, Cohn DH, Krakow D, Mottes M, Schwarze U, Chen D, Yang K, Kuslich C, Troendle J, Dalgleish R, Byers PH.Consortium for osteogenesis imperfecta mutations in the helical domain of type I collagen: regions rich in lethal mutations align with collagen binding sites for integrins and proteoglycans. Hum Mutat. 2007;28(3):209–221.doi:10.1002/ humu.20429. 28. Rossi V, Lee B, Marom R.Osteogenesis imperfecta: advancements in genetics and treatment. Curr Opin Pediatr. 2019;31(6):708–715.doi:10.1097/ MOP.0000000000000813. 29. Binh HD, Maasalu K, Dung VC, Ngoc CT, Hung TT, Nam TV, Nhan LN, Prans E, Reimann E, Zhytnik L, Kõks S, Märtson A.The clinical features of osteogenesis imperfecta in Vietnam. Int Orthop. 2017;41(1):21–29. 30. Lindahl K, Åström E, Rubin CJ, Grigelioniene G, Malmgren B, Ö L, Kindmark A.Genetic epidemiology, prevalence, and genotype–phenotype correlations in the Swedish population with osteogenesis imperfecta. Eur J Hum Genet. 2015;23(8):1042–1050.doi:10.1038/ ejhg.2015.81. 31. 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.doi:10.1038/ejhg.2009.242. 32. Sillence D, Butler B, Latham M, Barlow K.Natural history of blue sclerae in osteogenesis imperfecta. Am J Med Genet. 1993;45(2):183–186.doi:10.1002/ ajmg.1320450207. 33. Forlino A, Marini JC.Osteogenesis imperfecta. Lancet. 2016;387(10028):1657–1671. .doi:10.1016/S0140- 6736(15)00728-X. 34. Cabral WA, Chang W, Barnes AM, Weis M, Scott MA, Leikin S, Makareeva E, Kuznetsova NV, Rosenbaum KN, Tifft CJ, Bulas DI, Kozma C, Smith PA, Eyre DR, Marini JC.Prolyl 3-hydroxylase 1 deficiency causes a recessive metabolic bone disorder resembling lethal/severe osteogenesis imperfecta. Nat Genet. 2007;39(3):359–365.doi:10.1038/ng1968. 35. Marini JC, Cabral WA, Barnes AM, Chang W. Components of the collagen prolyl 3-hydroxylation complex are crucial for normal bone development. Cell Cycle. 2007;6(14):1675–1681.doi:10.4161/ cc.6.14.4474. 36. Santana A, Franzone JM, McGreal CM, Kruse RW, Bober MB. A moderate form of osteogenesis imper- fecta caused by compound heterozygous LEPRE1 mutations. Bone Rep. 2018;9:132–135.doi:10.1016/j. bonr.2018.09.002. 37. Alcausin MB, Briody J, Pacey V, Ault J, McQuade M, Bridge C, Engelbert RH, Sillence DO, Munns CF. Intravenous pamidronate treatment in children with moderate-to-severe osteogenesis imperfecta started under three years of age. Horm Res Paediatr. 2013;79 (6):333–340.doi:10.1159/000351374. 38. Rauch F, Travers R, Plotkin H, Glorieux FH.The effects of intravenous pamidronate on the bone tissue of chil- dren and adolescents with osteogenesis imperfecta. J Clin Invest. 2002;110(9):1293–1299.doi:10.1172/ JCI0215952. 39. Fujiwara I, Ogawa E, Igarashi Y, Ohba M, Asanuma A. Intravenous pamidronate treatment in osteogenesis imperfecta. Eur J Pediatr. 1998;157(3):261–262. 40. Dwan K, Phillipi CA, Steiner RD, Basel D. Bisphosphonate therapy for osteogenesis imperfecta. Cochrane Database Syst Rev. 2014;7:CD005088. 41. Letocha AD, Cintas HL, Troendle JF, Reynolds JC, Cann CE, Chernoff EJ, Hill SC, Gerber LH, Marini JC.Controlled trial of pamidronate in children with types III and IV osteogenesis imperfecta confirms vertebral gains but not short-term functional improvement. J Bone Miner Res. 2005;20(6):977–986. doi:10.1359/JBMR.050109. 42. Rauch F, Cornibert S, Cheung M, Glorieux FH.Long- bone changes after pamidronate discontinuation in chil- dren and adolescents with osteogenesis imperfecta. Bone. 2007;40(4):821–827.doi:10.1016/j.bone.2006.11.020. 43. Bains JS, Carter EM, Citron KP, Boskey AL, Shapiro JR, Steiner RD, Smith PA, Bober MB, Hart T, Cuthbertson D, Krischer J, Byers PH, Pepin M, Durigova M, Glorieux FH, Rauch F, Sliepka JM, Sutton VR, Lee B.Members of the BBD Consortium, Nagamani SC, Raggio CL. A Multicenter Observational Cohort Study to Evaluate the Effects of Bisphosphonate Exposure on Bone Mineral Density and Other Health Outcomes in Osteogenesis Imperfecta JBMR Plus. 2019;3(5):e10118. 44. Shi CG, Zhang Y, Yuan W.Efficacy of bisphosphonates on bone mineral density and fracture rate in patients with osteogenesis imperfecta: a systematic review and meta-analysis. Am J Ther. 2016;23(3):e894–904. doi:10.1097/MJT.0000000000000236. 45. Ralston SH, Gaston MS. Management of osteogenesis imperfecta. Front Endocrinol (Lausanne). 2020;10:924. doi:10.3389/fendo.2019.00924. 46. Phillipi CA, Remmington T, Steiner RD. Bisphosphonate therapy for osteogenesis imperfecta. Cochrane Database Syst Rev. 2008;(4):CD005088. DOI:10.1002/14651858.CD005088.pub2. 47. Hald JD, Evangelou E, Langdahl BL, Ralston SH. Bisphosphonates for the prevention of fractures in osteogenesis imperfecta: meta-analysis of placebo-controlled trials. J Bone Miner Res. 2015;30 (5):929–933.doi:10.1002/jbmr.2410. 48. Burr DB, Liu Z, Allen MR. Duration-dependent effects of clinically relevant oral alendronate doses on cortical bone toughness in beagle dogs. Bone. 2015;71:58–62. doi:10.1016/j.bone.2014.10.010. 49. Uveges TE, Kozloff KM, Ty JM, Ledgard F, Raggio CL, Gronowicz G, Goldstein SA, Marini JC.Alendronate treatment of the brtl osteogenesis imperfecta mouse improves femoral geometry and load response before fracture but decreases predicted material properties and has detrimental effects on osteoblasts and bone formation. J Bone Miner Res. 2009;24(5):849–859. doi:10.1359/jbmr.081238. 50. Ward LM, Rauch F, Whyte MP, D’Astous J, Gates PE, Grogan D, Lester EL, McCall RE, Pressly TA, Sanders JO, Smith PA, Steiner RD, Sullivan E, Tyerman G, Smith-Wright DL, Verbruggen N, Heyden N, Lombardi A, Glorieux FH. Alendronate for the treatment of pediatric osteogenesis imperfecta: a randomized placebo-controlled study. J Clin Endocrinol Metab. 2011;96(2):355–364.doi:10.1210/ jc.2010-0636.