Abstract
Background To describe a staged surgical protocol combining halo-pelvic traction (HPT) and posterior spinal fusion (PSF) for severe scoliosis in a patient with osteogenesis imperfecta (OI) type IV and to evaluate its outcomes. Given the paucity of population-level data on spinal orthoses in OI, this report highlights a tailored surgical approach for this high-risk population.
Case Presentation and Management A 16-year-old girl with OI type IV and progressive scoliosis underwent a 2-stage correction: (1) preoperative HPT for 3 months to reduce coronal deformity and optimize spinal alignment, followed by (2) PSF with all-pedicle-screw instrumentation. The staged protocol achieved successful deformity correction without neurological or implant-related complications. All pedicle screws were safely placed despite osteopenic bone. At follow-up, radiographic outcomes were maintained, and the patient reported improved posture and function. Minor surgical differences and literature review are highlighted for multimodal management.
Conclusion Progressive scoliosis in patients with OI can be effectively managed through structured, phased therapeutic programs, with the combined approach of HPT and PSF representing a significant surgical intervention strategy.
Clinical Relevance The clinical significance of this approach lies in transforming the management of a challenging rare disease—progressive scoliosis in osteogenesis imperfecta—from an empirical endeavor into a structured, systematic clinical pathway, while providing a validated technical combination for its most critical surgical intervention.
Level of Evidence 5.
Introduction
Osteogenesis imperfecta (OI) is a rare genetic disease in orthopedics, with a reported incidence of 1/15,000–20,000 in crowds.1 The main pathogenic factors are associated with variants in COL1A1 and COL1A2, as well as other mechanisms related to collagen biosynthesis, osteoblast differentiation, and bone mineralization.2 These individuals typically have slender bones and short stature due to poor growth patterns. Additional manifestations include blue sclera, dentinogenesis imperfecta, fragility of bones, and issues with hearing, lung function, and cardiovascular disorders.2,3 In 1979, Sillence et al first reported and detailed the discovery and defined types I to IV categories of OI.4 Subsequent research expanded the classification to 7 types based on clinical symptoms and genetic testing.5 New pathogenic mechanisms have been discovered, and rare types of OI are prompting investigators to re-evaluate established concepts such as collagen modification, folding, and cross-linking.6 Individuals with OI type IV typically experience recurrent fractures, osteoporosis, and some bone deformities, while having normal sclerae.7 In addition to common variables, we found blue sclerae and dentinogenesis imperfecta and identified a dominant genetic mutation in the COL1A1 gene. The severe form of OI is lethal before birth, while progressive OI may not be identified until adulthood. Differential diagnosis can typically be made through prenatal ultrasound and confirmed with various imaging techniques, genetic testing, and specific changes.8
Over the past 20 years, research on the treatment of OI has remained ongoing. While a complete cure is not currently achievable, the primary goals are to maximize function, minimize deformity and disability, ensure comfort, foster relative independence in daily activities, and promote social integration.9 The Marini JC program emphasizes the management of brittle bones by focusing on reducing fractures and enhancing overall function.10 Skeletal bone management involves monitoring bone density and fracture rates and reshaping vertebrae after compression fractures. Intravenous bisphosphonate therapy is the most commonly used method for treating bone fragility in children. Other therapies include denosumab, synthetic parathyroid hormone, and growth hormone.11,12 While popular belief suggests that bisphosphonates can increase bone density, a meta-analysis indicates that neither oral nor intravenous bisphosphonates significantly improve the clinical status of OI.13 Overall, there were no adverse outcomes. Treatment studies aim to develop personalized, genotype-based approaches like stem cell transplantation and genetic engineering to achieve breakthroughs for patients.14,15 Surgical intervention is necessary for fractured or deformed bones, requiring a multidisciplinary approach,16 which can be challenging.
Case Presentation
A 16-year-old girl was diagnosed with OI at an early age, and there was no indication that her family members were affected by the disease. Throughout her growth, she experienced a higher susceptibility to fractures and dislocations caused by trauma, resulting in multiple treatments for humeral and femoral fractures as well as joint dislocations, although precise details are unavailable. At the age of 13 years, the patient developed significant spinal deformity. Notable clinical features included a characteristic blue sclera, dentin hypoplasia, thin limbs, bone fragility, osteopenia, and loose interphalangeal ligaments; however, she did not have hearing loss (Figure 1). Whole-length EOS Imaging System radiography revealed thin cortical bone with lucent shadows, along with a double principal curve pattern in the spine, specifically an upper thoracic curvature toward the left and lower lumbar curvature toward the right, accompanied by thoracic kyphosis (Figure 2). A dual-energy x-ray absorptiometry scan indicated severe osteoporosis, and the patient had not received standardized antiosteoporotic treatment for this condition. Lung function was mildly impaired, while neurological assessments were normal. Genetic testing revealed a pathogenic mutation in the COL1A1 gene, specifically the c.3910C→T mutation on chromosome 17: 48263773. This mutation results in the premature termination of protein synthesis after the coding of glutamine, leading to disease manifestation. Genetic analysis confirmed a diagnosis of scoliosis associated with OI type IV. This study comprehensively delineates the implementation of a staged surgical protocol combining halo-pelvic traction (HPT) and posterior spinal fusion (PSF), along with integrated multidisciplinary management, for the treatment of severe thoracic hyperkyphosis and spinal stiffness.
Characteristic manifestation. Preoperative lateral (A) and back (B) images showing spinal deformity. (C) Blue sclera. (D) Dentin hypoplasia. (E) Thin lower limbs. (F) Loose interphalangeal ligaments.
Preoperative whole-length EOS imaging of patient in anteroposterior (A) and lateral (B) views showing scoliosis and kyphosis, thin cortical bone, and osteoporosis.
Surgery and Imaging
When we first evaluated the patient’s entire spine with EOS imaging, it revealed the presence of scoliosis and kyphosis. The Cobb angle for the thoracic curve, measured from T4 to T12, was 104°, while the Cobb angle for the lumbar curve, spanning T12 to L4, was 66°. In the sagittal plane, the thoracic kyphosis from T4 to T12 measured 67° (Figure 2). Notably, the vertebrae in the upper curve showed a tendency to fuse and were associated with both coronal and sagittal imbalances. Given the patient’s severe osteoporosis, ligamentous laxity, and propensity for rough movements, concerns were raised about the risk of vertebral fractures and serious complications following orthosis application. Evidence suggested that a staged surgical approach be adopted to gradually improve spinal alignment and stability. Initial HPT was employed to reduce deformity, enhance biomechanical stability, facilitate tissue release, address some of the angular deformities, and minimize risks while optimizing long-term outcomes. This strategy prioritizes safety and controlled progression, addressing the patient’s complex spinal pathology effectively.
Traction Surgery
The patient was placed supine with the head suspended and the waist elevated. A skull ring equipped with 4 pairs of titanium pins was fixed symmetrically above the ears, aligned with the body axis. The anterior pairs were positioned at the superior lateral border of the eyebrow arch, and the posterior pairs were placed on the posterolateral aspect of the occipital bone. Under fluoroscopic guidance, the pins were securely inserted into the outer skull table without intracranial penetration. A pelvic traction pin was inserted 4 cm posterior to the anterior superior iliac spine and guided through the posterior superior iliac spine from anterior to posterior, ensuring symmetrical length on both sides. Fluoroscopy confirmed no pelvic penetration or fracture. Four longitudinal connecting rods were then attached to provide balanced traction in both coronal and sagittal planes (Figure 3A). Distraction commences following pain resolution, initially at a rate of 1 to 2 cm per day for 1 week, then reduced to 2 to 5 mm every other day until the tolerance limit is reached.17 Gradual force augmentation is essential to allow progressive adaptation of bone, muscle, and ligamentous tissues, thereby mitigating the risk of complications such as nerve traction injury or soft tissue damage. In the later stage of traction, she slept peacefully on the sponge bed that was preconcave in advance.
Illustration of the instrument. (A) Halo-pelvic traction. (B) 3D-printed models. (C) Prone on the sponge bed with a ring in the operation.
During the traction period, the patient demonstrated excellent tolerance and compliance with rehabilitation exercises, which contributed to an uneventful surgical procedure without any adverse events. Nevertheless, the intervention was associated with a transient decline in quality of life. Subsequent follow-up monitoring revealed an approximately 4-cm increase in patient height.
Fusion Surgery
After 3 months of HPT, the whole spine EOS anterior-to-lateral view showed a reduction in the primary thoracic angle from 104° to 84° (19% correction rate) and the secondary lumbar angle from 66° to 58° (12% correction rate). The kyphosis curve decreased from 67° to 46° (31% correction rate) (Figure 4; Table). The patient underwent a PSF instrumented correction from T2 to L5 after being readmitted to the hospital. The patient was in the prone position on the operating bed with the traction ring and supported by a sponge bed (Figure 3). After the segments were determined by fluoroscopy before surgery, all pedicle screws were placed using the manual pedicle screw placement technique. The posterior metal structure consists of 5.0 mm to 6.5 mm pedicle screws and 5.5 mm rods (titanium). The articular process and transverse process were exposed after the soft tissues were separated, and the base and articular surface of the superior articular process were marked. The projection location of the pedicle was determined, and a mouth opener was used to create a 4- to 5-mm notch in the cortex. A metal cone was then used to manually form the pedicle channel. A flexible metal ball detector was tested separately on 5 different bone boundaries (1 base and 4 walls). Using a tap smaller than the pedicle screw, the nail path was tapped out, and the pedicle screw was placed along the channel after verification. The size of the bone wall and screw was based on the surgeon’s experience and judgment. We aimed to place screws with the largest diameter possible to ensure a better purchase power of bone at each level. The entire procedure was monitored by neuroelectrophysiology, and each pedicle screw was implanted with electrical stimulation to ensure safety. Following verification of screw placement accuracy via C-arm fluoroscopy, the rod—pre-contoured to match the patient’s physiological curvature—was first engaged on the concave side. Subsequently, the rod was positioned on the convex side, and gradual corrective pressure was applied to reduce the deformity. After achieving the desired goal, all the spine processes in the rear were removed using bone rongeurs, and autologous bone grafts were performed along with all the previous bones, in addition to artificial bones. Last, we removed the patient’s HPT device in accordance with routine.
After the first operation with halo-pelvic traction (HPT) of the whole spine. EOS imaging, anteroposterior (A) and lateral (B). After HPT for 3 months, deformity has been slightly corrected, as shown in anteroposterior (C) lateral (D) imaging.
Cobb angle measures over time at baseline, after traction surgery, after fusion surgery, and follow-up.
To reduce the risk of implant withdrawal and minimize force exerted during orthopedic surgery, we utilized an ultrasonic osteotome to perform a resection of the lamina and articular process at the lower end on a small scale while avoiding the apex cone. This technique is based on the principle of posterior column osteotomy. The shape of the resection was designed like a transverse “V,” resembling a standing bow tie. This configuration promotes sufficient flexibility and allows for better correction during the rod placement process. Additionally, we removed approximately 4 cm of 2 ribs from the right costal joint.
A 3D-printed model for the patient was a good reference for intraoperative nail placement and orthosis before the operation (Figure 3). The total operative bleeding was approximately 1250 mL, and the procedure duration was 450 minutes. During the operation, both autologous and allogeneic blood transfusions were administered.
Outcome and Follow-Up
The patient was able to ambulate independently for short periods postoperatively but complained of lower back pain. The imaging examinations confirmed the stability of all internal fixations, revealing no vertebral fractures or neurological symptoms. The EOS imaging showed satisfactory correction results (Figure 5). A tailor-made thoracolumbar brace was given to the patient, who immediately began the home rehabilitation phase.
EOS imaging after the second operation of the whole spine with metal implant in anteroposterior (A) and lateral (B). views. Anteroposterior (C) and lateral (D) views at 3-mo follow-up. Anteroposterior (E) and lateral (F) views at 6-mo follow-up. Postural assessment parameters showed no loss of correction on the coronal and sagittal planes.
The patient demonstrated an excellent recovery, with markedly increased daily activity levels. At 3 months of follow-up, EOS imaging revealed a Cobb angles 30° thoracic and the 32° lumbar, with the kyphosis curve decreased to 18° (Figure 5; Table). After reviewing the EOS and completing the necessary imaging at subsequent follow-up, we did not find any signs of significant prolapse or loosening of the internal fixation, and there was no loss of the correction degree. The patient and her family were informed that relevant data would be submitted for publication, and written consent was obtained.
Discussion
OI patients present with ligamentous laxity and bone fragility. Abnormalities can occur in any part of the spine (craniocervical junction, scoliosis, kyphosis, osteoporotic fractures, degenerative disc disease, spondylolisthesis, and spondylolysis); these abnormalities require surgery to resolve.2,18 The incidence of scoliosis in OI is reported to be between 39% and 80%,18,19 and curve progression varies with growth (1°/y in type I, 4°/y in type IV, and 6°/y in type III).20 Lubicky suggested that surgeons ought to plan an effective intervention sooner rather than later when facing progressive or symptomatic deformity.21 For scoliosis angles >50°, surgery should be considered to slow the deterioration of lung function.22 Pan et al reported using a 3-rod all-pedicle screw fixation technique to achieve strong fusion and fixation of the patient.23 Ono et al treated a case of early-onset scoliosis by adjusting the growth rod several times, which led to a good outcome.24 Jones et al used conventional pedicle hooks and sublaminar wires to treat a case of type IV scoliosis that did not lead to hyperplastic callus.25 Chehrassan et al examined 6 surgically treated OI patients with pedicle screws and hooks and emphasized that Ponte osteotomy may improve the correction and reduce the necessary force.26 There have also been reports of the first use of all pedicle screw constructs with cement augmentation of screws to prevent screw failure at the proximal and distal anchor points.27 Gurel et al conducted a case series that reviewed the safety of cementless PSF for treatment, and the halo gravity traction (HGT) scheme was described.28 A preliminary study showed that preoperative halo-traction was a friendly adjunct to increase the chance of fusion success.29 Janus et al reported 20 patients in whom partial correction of the spine was obtained using an HGT device preoperatively,30 which seemed to indicate that operative stabilization is possible. The management of spinal deformities in OI remains formidable. Despite reports of successful outcomes with personalized surgical approaches, compromised bone quality secondary to severe osteoporosis and ligamentous laxity presents substantial obstacles. These often render surgery prohibitive or are associated with a high incidence of device-related complications and instrumentation failure,29,31 frequently resulting in an inability to curtail the progression of spinal pathology.
In terms of preoperative correction, one study demonstrated that HPT is more effective than HGT in improving spinal deformities, reducing osteotomy grades, and decreasing the number of fusion segments.17 HPT employs fixation at the cranial and pelvic regions, providing more direct and controllable corrective forces. This technique facilitates improvement in both coronal and sagittal plane deformities, increases spinal length, and avoids the need for high-grade osteotomies. To our knowledge, the present report represents the first documented application of this technique in OI. The construct provided precise 3-dimensional adjustability, and the applied traction effectively induced gradual spinal straightening, demonstrating its feasibility for correcting severe spinal deformities in OI. The most important factor is that pedicle screws were used in all vertebral bodies, which provides a strong purchase of bone. The pedicle screw has higher independence of bone mineral density and provides greater resistance to tangential loading than the hook. This approach eliminates the need for repeated growing rod revision surgeries and avoids systemic toxicity from high-volume cement augmentation. It can be applied to fragile, flimsy, and feeble vertebrae. However, the placement of all pedicle screws makes the procedure complicated, which increases the operation time and blood loss. Moreover, we also took the risk of using posterior column osteotomy, only on a small scale, limited to the lower thoracic vertebrae, not in the parietal cone. When strategically performed for osteotomy, it offered considerable benefits in correcting curvature while minimizing risks linked with implant failure. In managing instrumented spinal fusion in our patient, autograft bone and allograft bone are used as bone implants. Although pharmacotherapy for bone density enhancement is widely recommended in the literature, detailed surgical protocols and their critical roles remain poorly documented. Our management strategy prioritizes critical complications in severe spinal manifestations of OI while maintaining anti-osteoporosis therapy as the cornerstone, supplemented by longitudinal postoperative surveillance.
The experience of our case is staged foundation traction and fusion. Surgical treatment of scoliosis in patients affected by OI is challenging and risky, with peri- and postoperative complications. In our case, various strategies have been employed to mitigate implant failure. It is essential to gradual correction without attempting to overcorrect. The management of OI requires a multidisciplinary approach, and treatment options include conservative measures, medication treatment, orthopedic surgery, and rehabilitation treatment. Especially for growing children, the spine has the potential to grow. Perhaps a visionary target in the future is an individualized, specific-mutations approach.
Conclusion
Scoliosis and kyphosis in a patient with OI were effectively managed through a staged approach that involved foundation traction and fusion, with the application of HPT for extended periods. Nevertheless, the population-level efficacy of this combined surgical approach requires further validation. We believe that personalized and progressive surgical strategies can better treat patients with these rare conditions.
Footnotes
Funding This work was supported by the National Natural Science Foundation of China (Grant No:82460260), Yunnan Province High-Level Talent Support Program (Grant No:XDYC-YLWS-2024-0072), Young and Middle-Aged Academic and Technical Leadership Reserve (Grant No:202405AC350059), Key Research Project of Yunnan Provincial Science and Technology (Grant No:202402AG050001-02), and Yunnan Spinal Cord Disease Clinical Medical Center (ZX2022000101-2024JSKFKT-02).
Declaration of Conflicting Interests The authors report no conflicts of interest in this work.
Patient Consent Informed (written and verbal) consent was obtained from the legal guardians for publication of this manuscript and the accompanying images.
- This manuscript is generously published free of charge by ISASS, the International Society for the Advancement of Spine Surgery. Copyright © 2025 ISASS. To see more or order reprints or permissions, see http://ijssurgery.com.
References
- 1.↵Baljet B . Aspects of the history of osteogenesis imperfecta (Vrolik’s syndrome). Ann Anat. 2002;184(1):1–7. 10.1016/S0940-9602(02)80023-1
- 2.↵Marom R , Rabenhorst BM , Morello R . Osteogenesis imperfecta: an update on clinical features and therapies. Eur J Endocrinol. 2020;183(4):R95–R106. 10.1530/EJE-20-0299
- 3.↵Thomas IH , DiMeglio LA . Advances in the classification and treatment of osteogenesis imperfecta. Curr Osteoporos Rep. 2016;14(1):1–9. 10.1007/s11914-106-0299-y
- 4.↵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
- 5.↵Rauch F , Glorieux FH . Osteogenesis imperfecta. Lancet. 2004;363(9418):1377–1385. 10.1016/S0140-6736(04)16051-0
- 6.↵Forlino A , Marini JC . Osteogenesis imperfecta. Lancet. 2016;387(10028):1657–1671. 10.1016/S0140-6736(15)00728-X
- 7.↵Van Dijk FS , Sillence DO . Osteogenesis imperfecta: clinical diagnosis, nomenclature and severity assessment. Am J Med Genet A. 2014;164A(6):1470–1481. 10.1002/ajmg.a.36545
- 8.↵Deguchi M , Tsuji S , Katsura D , Kasahara K , Kimura F , Murakami T . Current overview of osteogenesis imperfecta. Medicina (Kaunas). 2021;57(5):464. 10.3390/medicina57050464
- 9.↵Kocher MS , Shapiro F . Osteogenesis imperfecta. J Am Acad Orthop Surg. 1998;6(4):225–236. 10.5435/00124635-199807000-00004
- 10.↵Marini JC . Osteogenesis imperfecta--managing brittle bones. N Engl J Med. 1998;339(14):986–987. 10.1056/NEJM199810013391408
- 11.↵Roberts TT , Cepela DJ , Uhl RL , Lozman J . Orthopaedic considerations for the adult with osteogenesis imperfecta. J Am Acad Orthop Surg. 2016;24(5):298–308. 10.5435/JAAOS-D-15-00275
- 12.↵Palomo T , Vilaça T , Lazaretti-Castro M . Osteogenesis imperfecta: diagnosis and treatment. Curr Opin Endocrinol Diabetes Obes. 2017;24(6):381–388. 10.1097/MED.0000000000000367
- 13.↵Dwan K , Phillipi CA , Steiner RD , Basel D . Bisphosphonate therapy for osteogenesis imperfecta. Cochrane Database Syst Rev. 2014;(7):CD005088. 10.1002/14651858.CD005088.pub3
- 14.↵Rossi V , Lee B , Marom R . Osteogenesis imperfecta: advancements in genetics and treatment. Curr Opin Pediatr. 2019;31(6):708–715. 10.1097/MOP.0000000000000813
- 15.↵Botor M , Fus-Kujawa A , Uroczynska M , et al . Osteogenesis imperfecta: current and prospective therapies. Biomolecules. 2021;11(10):1493. 10.3390/biom11101493
- 16.↵Hidalgo Perea S , Green DW . Osteogenesis imperfecta: treatment and surgical management. Curr Opin Pediatr. 2021;33(1):74–78. 10.1097/MOP.0000000000000968
- 17.↵Zhou L , Wang J , Yang H , Zhang Y , Wang Y , Hai Y . Efficacy of halo pelvic traction versus halo gravity traction in treating severe rigid spinal deformities: a matched retrospective study. Global Spine J. 2025:21925682251348670. 10.1177/21925682251348670
- 18.↵Puvanesarajah V , Sponseller PD . Scoliosis and kyphosis in osteogenesis imperfecta. Osteogenesis Imperfecta. San Diego: Academic Press; 2014:401–406. 10.1016/B978-0-12-397165-4.00043-5
- 19.↵Renshaw TS , Cook RS , Albright JA . Scoliosis in osteogenesis imperfecta. Clin Orthop Relat Res. 1979;(145):163–167.
- 20.↵Anissipour AK , Hammerberg KW , Caudill A , et al . Behavior of scoliosis during growth in children with osteogenesis imperfecta. J Bone Joint Surg Am. 2014;96(3):237–243. 10.2106/JBJS.L.01596
- 21.↵Lubicky JP . The spine in osteogenesis imperfecta. Spine Deform. 2012;1(1):124–132. 10.1016/j.jspd.2012.04.006
- 22.↵Widmann RF , Bitan FD , Laplaza FJ , Burke SW , DiMaio MF , Schneider R . Spinal deformity, pulmonary compromise, and quality of life in osteogenesis imperfecta. Spine (Phila Pa 1986). 1999;24(16):1673. 10.1097/00007632-199908150-00008
- 23.↵Pan C-H , Ma S-C , Wu C-T , Chen P-Q . All pedicle screw fixation technique in correcting severe kyphoscoliosis in an osteogenesis imperfecta patient: a case report. J Spinal Disord Tech. 2006;19(5):368–372. 10.1097/01.bsd.0000208253.06706.42
- 24.↵Ono Y , Miyakoshi N , Hongo M , et al . Growing rod surgery for early-onset scoliosis in an osteogenesis imperfecta patient. World Neurosurg. 2020;144:178–183. 10.1016/j.wneu.2020.08.165
- 25.↵Jones M , Breakwell L , Cole A , Arundel P , Bishop N . Type V osteogenesis imperfecta undergoing surgical correction for scoliosis. Eur Spine J. 2018;27(9):2079–2084. 10.1007/s00586-018-5465-8
- 26.↵Chehrassan M , Shakeri M , Nikouei F , Jafari B , Mahabadi EA , Ghandhari H . Surgical treatment of scoliosis in patients with osteogenesis imperfecta: a single institution case series. Musculoskelet Surg. 2024;108(4):437–441. 10.1007/s12306-024-00842-0
- 27.↵Yilmaz G , Hwang S , Oto M , et al . Surgical treatment of scoliosis in osteogenesis imperfecta with cement-augmented pedicle screw instrumentation. J Spinal Disord Tech. 2014;27(3):174–180. 10.1097/BSD.0b013e3182624b76
- 28.↵Gurel R , Ovadia D , Lebel DE , Zeitlin L , Sigal A . Cementless posterior spinal fusion for the treatment of OI patients with severe spine deformity-a case series. Eur Spine J. 2022;31(6):1566–1572. 10.1007/s00586-022-07179-x
- 29.↵Badin D , Mun F , Akbarnia BA , Perez-Grueso F , Sponseller PD , Pediatric Spine Study Group . Outcomes of growth-friendly instrumentation in osteogenesis imperfecta: a preliminary report. J Pediatr Orthop. 2023;43(6):e458–e464. 10.1097/BPO.0000000000002405
- 30.↵Janus GJ , Finidori G , Engelbert RH , Pouliquen M , Pruijs JE . Operative treatment of severe scoliosis in osteogenesis imperfecta: results of 20 patients after halo traction and posterior spondylodesis with instrumentation. Eur Spine J. 2000;9(6):486–491. 10.1007/s005860000165
- 31.↵Yong-Hing K , MacEwen GD . Scoliosis associated with osteogenesis imperfecta. J Bone Joint Surg Br. 1982;64(1):36–43. 10.1302/0301-620X.64B1.7068718
