Osteogenesis Imperfecta: Differentiating Clinical Presentation from Non-Accidental Injury

 ABSTRACT:

This paper will focus on differentiating the pathologic fractures of osteogenesis imperfecta (OI), a bone fragility disorder that manifests in childhood, from the fractures typically sustained due to child abuse and neglect. Both phenomena can present similarly, with poorly explained, atypical fractures for developmental age. OI also presents with a multitude of physical findings outside the skeletal system involving the eyes, teeth and skin that can be clues to the diagnosis. As part of management, OI requires early bisphosphonate therapy to prevent lasting deformity and spinal complications into adulthood. By exploring the radiographic evidence in conjunction with clinical signs and symptoms of osteogenesis imperfecta, this review of the literature will focus on the workup necessary to diagnose and distinguish this condition from others without relying on genetic testing.

INTRODUCTION:

Osteogenesis imperfecta (OI) is a rare, hereditary disorder of bone fragility affecting only 1/10,000 individuals worldwide, that typically presents in childhood with spontaneous fractures, or fractures sustained due to seemingly insignificant trauma.¹ Non-accidental injury (NAI) in childhood is a phenomenon that can present similarly, and due to its high morbidity, is necessary to rule out when making the diagnosis of OI. Although OI can be confirmed by genetic testing, this can be both inconclusive and cost prohibitive, so it is important to be able to make a clinical diagnosis. In order to effectively discern OI from instances of NAI, extramedullary findings as well as orthopedic data are imperative to review. More specifically, because both OI and NAI can present with anomalous orthopedic manifestations, analyzing the type, location and quantity of fractures can help clinicians more aptly recognize OI versus abusive pathology.

Osteogenesis imperfecta is the most common inherited disorder of bone fragility in the pediatric population. Its characteristic predisposition to fractures is caused by autosomal dominant mutations in COL1A1 or COL1A2 genes in 90% of cases, although up to 17 new gene mutations have been discovered over the last decade.² These mutations cause abnormal collagen cross-linking, and alter the triple helical structure of type I collagen via a glycine substitution. This results in either reduced production of normal collagen, or production of structurally abnormal collagen. Subsequently, periosteal osteoblasts cannot form enough osteoid, resulting in brittle bone leading to frequent fracture, and improper remodeling leading to progressive bowing and deformity of the extremities.^2,3

Due to the multiple implicated gene mutations and resulting phenotypic heterogeneity in disease presentation, a classification scheme was proposed by Sillence et al. in 1979 and is still in use to categorize OI today. OI Type 1 (OI-1) is the mildest and most common form of the disease, in which patients have few fractures and minimal bone deformity. OI Type 2 (OI-2) is the most severe form, as it is lethal in the neonatal period and presents with significant bone fragility. OI Type 3 (OI-3) is also severe but not immediately lethal, manifesting with multiple fractures, pervasive deformity and short stature. OI Type 4 (OI-4) is described as moderate with vast clinical variability. OI Type 5 (OI-5) is also a moderate disease state but with several distinct clinical and radiological findings, including calcification of the interosseous membrane between long bones, hyperplastic callus formation, radial head dislocations, and the absence of dental manifestations.^1,4 Because of the widespread nature of genetic screening in pregnancy, incidence of the lethal forms of OI have dropped to less than 5% of OI births, meaning that the majority of infants with OI have the milder subsets with less apparent pathologic features.¹ It is these milder forms of OI that are the most likely to be mistaken for non-accidental trauma in an otherwise healthy child.

In addition to frequent fractures, other physical signs of the disease can inform diagnosis. Type I collagen is not only found in bone, but is a structural protein integral to the development of skin, tendon, dentin and sclera, leading to phenotypic findings outside the skeletal system that can provide clues to a diagnosis of OI.² Blue sclera is one of the most common physical findings in OI patients, observed in upward of 90% of cases, but can also be a normal variant in healthy children. Wormian bones, which are extra bones found between the sutures and fontanelles of the skull on imaging, are seen in approximately 30% of OI patients. Dentinogenesis imperfecta, a third anomalous finding impacting up to 50% of people with OI alters dentin formation and results in amber-colored, translucent primary and secondary teeth.⁴ Other non-orthopedic findings of OI include hearing loss, joint hypermobility, short stature, thin skin predisposed to to subcutaneous hemorrhage, cardiovascular complications, triangular face and and basilar skull invagination.^2,4 Although these findings are not specific to OI and can be present in other disease states or in the normal pediatric population, they can be helpful when combined with orthopedic data to solidify a diagnosis.

 Blue sclera     Dentinogenesis imperfecta ⁵

Among the differential for OI and frequent fractures includes hypophosphatasia, disorders of copper metabolism, rickets and non-accidental injury.⁶ Of these diagnostic possibilities, NAI can be the most fatal, yet the most preventable if effectively picked up in a healthcare setting. The incidence of non-accidental injury in which fractures are incurred is 24/10,000 pediatric patients under three years of age.⁷ Because abuse is a more common picture than OI, mild forms of OI without obvious extramedullary findings may get misdiagnosed as NAI. Therefore, the clinician must take an in depth look at fracture pattern in addition to the mere presence of multiple fractures. Both OI and NAI present in early childhood with an injury pattern incongruous with history and developmental age, but closely analyzing the type and location of fracture can prevent misdiagnosis.

First, forming an understanding of how fractures classically present in the pediatric population can aid in the identification of fractures that are pathologic. Peddada et al. looked at routine pediatric fractures in comparison with fractures sustained due to OI, and demonstrated that patients with OI were often younger at the time of fracture. Their research also illustrated that OI patients had more oblique, transverse, diaphyseal and bilateral long bone fractures than their healthy pediatric counterparts, who had more buckle, metaphyseal and physeal fractures.⁸               In patients with normal bone composition, the diaphysis is the strongest area, which explains why healthy children did not exhibit diaphyseal fractures. Conversely, the physis is the weakest area of normal pediatric bone, as it is not yet calcified, which explains the prevalence of growth plate injuries in healthy children. Buckle fractures are also expected in healthy children, as pediatric bone is soft and pliable, lending itself to incomplete breaks. Additionally, these children did not have transverse or oblique fractures, which are caused by either tensile or shearing forces, respectively. Normal collagen provides tensile strength, but in instances like OI where collagen is abnormal or diminished, tensile stress via bending from even minimal trauma can lead to transverse fractures. Similarly, bones with deficient collagen cannot withstand shearing forces.⁸ Furthermore, it is notable that on skeletal survey, patients with OI often have osteopenia and multiple thoracolumbar compression fractures, which are rare in otherwise healthy children and those being physically abused.⁷ Having a baseline understanding of pediatric bone composition and typical injury patterns in healthy kids can then prompt the identification of pathologic variants that fall outside of this spectrum.

Children who have sustained abuse also fall outside of this spectrum, as their injuries differ in location and severity. They present with fracture patterns that are reflective of squeezing, shaking or blunt force trauma that would be impossible for a non-ambulatory infant to self-inflict. Clinical geneticist Elaine Pereira reports a 95% positive predictive value of abuse in children with symmetric, posteromedial rib fractures. This injury pattern is rarely seen in OI and is typically caused by picking up and forcefully squeezing a child at the ribcage.⁷ Metaphyseal corner fractures, complex fractures of the skull, sternum, scapula and vertebral spinous processes, as well as  transverse fractures of multiple digits in infants and young children also implicate forceful trauma as the mechanism of injury.^1,4,7 The fracture types associated with OI are distinctly caused by structural weakness, whereas the injuries associated with NAI require applied force as the mechanism of injury.

Additional research looking at fracture pattern at the time of OI diagnosis continues to differentiate OI from other conditions it may mimic. In a retrospective chart review conducted at The Shriner’s Hospital for Children in Houston, Texas, extremity fractures were found to be the most prevalent when OI was initially diagnosed. Rib fractures were also identified within this patient subset, but had a bimodal distribution of appearance. Patients with OI either had rib fractures prenatally and directly after birth, or as ambulatory children, but never in infancy. This is explained by birth trauma and activity level while learning to walk, and continues to elucidate that infantile rib fractures cannot be self inflicted, and therefore are almost always due to abuse. Another similar yet important finding within this same study was that multiple concurrent fractures are rare in OI outside the perinatal period where birth trauma is implicated. Upward of three unexplained fractures in children less than one year of age is almost always thought to be due to abuse.⁹ Looking at quantity and type of fracture, as well as the age at which the fracture was sustained, can be impactful in differentiating OI from non-accidental injury.

The value of a clinical diagnosis is supported by the lack of utility found in genetic testing. Genetic testing is often thought of as a way to a definitive diagnosis, however, testing for only COL1A1 and COL1A2 does not capture the entire population with OI, since de novo mutations are possible. In a retrospective study of genetic testing to rule out OI in cases of abuse, Zarate et al. found that when non-accidental injury was foremost on the differential, genetic testing for COL1A1 and COL1A2 was universally negative. Furthermore, the patients in this sample who did have COL1A1 or COL1A2 mutations also had clinical signs consistent with a diagnosis of OI. However, one patient had a mixed picture, which further complicates the diagnostic workup. This particular patient was hospitalized for fractures typically suggestive of abuse, including fractures of the femoral metaphysis, tibial metaphysis, scapula, and multiple ribs, yet she was found to have a COL1A2 variant.⁶ In cases like this, diagnosis is exceedingly challenging, and genetic testing proves necessary. Overall, genetic testing is merited only if there are concurrent clinical features of OI or there is a reason to believe NAI and OI exist simultaneously in the same patient.

The goals of treatment for patients with OI are to reduce fracture occurrence and lasting bone deformity, to increase mobility, and to bolster bone mass. For patients with mild OI, management may involve only acute fracture care.¹⁰ Bisphosphonates, however, have become the mainstay of treatment. While they are considered a supportive therapy and not a cure, bisphosphonates are profound in managing complications in growing children. They have been found to prevent long bone fractures, and to entirely reshape vertebral compression fractures.¹⁰ These drugs work by inhibiting osteoclast activity and thus increasing bone density. They can be administered orally or intravenously, with varying effects. A randomized study conducted by Lv F, et al. compared outcomes in patients receiving a once-yearly infusion of zoledronic acid with those receiving weekly oral Alendronate. Both treatment regimens worked comparably to increase bone density, but intravenous zoledronic acid was found to have a greater effect on reducing fracture frequency.¹¹ Typically, intravenous bisphosphonates are reserved for children with vertebral compression fractures or those with two or more long bone fractures per year, whereas oral bisphosphonates are utilized in mild to moderate cases of OI in children without compression fractures.¹²

Other spinal complications and long bone deformities require management outside of bisphosphonate therapy. Though bisphosphonates affect bone composition, they do not alleviate deformity, such as kyphosis, scoliosis, spondylosis, spondylolisthesis, coxa vara, and leg length discrepancies, which often arise in OI patients. These conditions are treated with a combination of exercise, bracing and surgical techniques including intramedullary rodding and epiphysiodesis, among others.¹³ Bisphosphonates in some cases, have actually been demonstrated to slow osteotomy healing after these procedures.¹⁰ Therefore, a multidisciplinary approach involving more than just pharmacotherapy alone should be taken to treating OI.

Both diagnosis and treatment of OI have been thoroughly researched, however, there are still some gaps in the existing literature with regard to the full history surrounding any given fracture. Because the mechanism is inconsistent with the presentation in OI and abuse, additional research is needed that not only contrasts type, location and quantity of fractures, but also includes data on how the injury came to be. Documentation of exact mechanism of injury as explained by the patient and family at the time of presentation to a healthcare facility could further explicate trends in the manifestations of OI versus NAI that could then be used diagnostically. Another future direction to explore in terms of treatment strategy for OI is to look at diet and exercise to see if lifestyle modifications make a statistically significant difference in bone composition in OI patients. Encouraging weight bearing exercise and frequent physical activity, as well as increasing consumption of calcium-rich foods could serve as an adjunct to the current standard of treatment for increasing bone mass.

Because the management of OI and NAI differs so vastly, early detection of both problems is imperative. Early orthopedic intervention and bisphosphonate treatment can curb lasting complications in OI, just as Child Protective Services involvement can stop the continued cycle of abusive injuries in NAI. Although genetic testing is available to pinpoint specific gene mutations causing OI, it is not necessary when physical exam findings, historical clues, and radiographic evidence are used appropriately. Based on the existing literature, it can be concluded that fracture characteristics at time of presentation play a role in differentiating between disease state and intentional harm to healthy children. Fractures demonstrating inherent structural weakness, such as mid-shaft tibial fractures, are indicative of metabolic bone disease. Conversely, fractures revealing force unlikely to be generated by a child, such as multiple posterior rib fractures, are suggestive of maltreatment. When mechanism of injury does not match the incurred trauma, clinicians should look to radiographic fracture patterns to distinguish OI from NAI and subsequently forward the timely recognition and treatment of disease.

References

 1. Pepin MG, Byers PH. What every clinical geneticist should know about testing for osteogenesis imperfecta in suspected child abuse cases. Am J Med Genet C Semin Med Genet. 2015;169(4):307-313.

2. Souder C. Osteogenesis Imperfecta. 2018; https://www.orthobullets.com/pediatrics/ 4102/osteogenesis-imperfecta. Accessed June 27, 2018.

3. Osteogenesis imperfecta. 2017. http://www.dynamed.com/topics/dmp~AN~T116818/ Osteogenesis-imperfecta – History-and-Physical. Accessed March 24, 2018.

4. Brizola E, Zambrano MB, Pinheiro BS, Vanz AP, Felix TM. Clinical Features and Pattern of Fractures at the Time of Diagnosis of Osteogenesis Imperfecta in Children. Rev Paul Pediatr. 2017;35(2):171-177.

5. UpToDate.; 2018. Available at: https://www.uptodate.com/contents/osteogenesis-imperfecta-clinical-features-and-diagnosis#H13. Accessed October 8, 2018.

6. Zarate YA, Clingenpeel R, Sellars EA, et al. COL1A1 and COL1A2 sequencing results in cohort of patients undergoing evaluation for potential child abuse. Am J Med Genet A. 2016;170(7):1858-1862.

7. Pereira EM. Clinical perspectives on osteogenesis imperfecta versus non-accidental injury. Am J Med Genet C Semin Med Genet. 2015;169(4):302-306.

8. Peddada KV, Sullivan BT, Margalit A, Sponseller PD. Fracture Patterns Differ Between Osteogenesis Imperfecta and Routine Pediatric Fractures. J Pediatr Orthop. 2018;38(4):e207-e212.

9. Greeley CS, Donaruma-Kwoh M, Vettimattam M, Lobo C, Williard C, Mazur L. Fractures at diagnosis in infants and children with osteogenesis imperfecta. J Pediatr Orthop. 2013;33(1):32-36.

10. Trejo P, Rauch F. Osteogenesis imperfecta in children and adolescents-new developments in diagnosis and treatment. Osteoporos Int. 2016;27(12):3427-3437.

11. Lv F, Liu Y, Xu X, et al. Zoledronic Acid Versus Alendronate in the Treatment of Children with Osteogenesis Imperfecta: A 2-Year Clinical Study. Endocr Pract. 2018;24(2):179-188.

12. Simm PJ, Biggin A, Zacharin MR, et al. Consensus guidelines on the use of bisphosphonate therapy in children and adolescents. J Paediatr Child Health. 2018;54(3):223-233.

13. Marr C, Seasman A, Bishop N. Managing the patient with osteogenesis imperfecta: a multidisciplinary approach. J