Clinically and genetically heterogeneous disorder manifested by bone fragility and low bone mass.
Seven distinct subtypes exist. Severity is as follows: type II > type III > types IV = V = VI = VII > type I.
Most cases that present prenatally are types II or III. Only 10% of fetuses with type I have fractures in utero.
Other findings include blue sclerae, abnormal teeth, joint hyperlaxity, adult-onset hearing loss, and normal intelligence.
Prenatal sonographic findings include long bone fractures with callus formation, limb shortening, poor mineralization of the skull, and bent femurs.
Differential diagnosis includes campomelic dysplasia, hypophosphatasia, and achondrogenesis.
In 90% of cases there is a mutation in one of the genes that codes for type I procollagen, COL1A1 or COL1A2.
Most cases are dominantly inherited. If parents are asymptomatic there is a 7% recurrence risk due to the surprisingly high incidence of gonadal mosaicism.
Osteogenesis imperfecta is a clinically and genetically heterogeneous disorder of connective tissue, manifested by bone fragility and low bone mass. Affected patients have blue sclerae, hearing abnormalities, defective dentition, hyperlaxity of the joints, and normal intelligence (Brons et al., 1988). The majority of affected individuals are heterozygous for mutations of the COL1A1 or COL1A2 gene, which alters the structure of type I procollagen (Cole and Dalgleish, 1995).
Osteogenesis imperfecta was originally classified into four clinically distinct disorders that were first delineated by Sillence et al. (1979), and modified by Rauch and Glorieux (2004) (Table 91-1). Type I is the common mild form, type II is the perinatal lethal form, type III is the severe form, and type IV is the moderately clinically severe form (Cole and Dalgleish, 1995). More recently, an additional three types (V, VI, and VII) have been described (Rauch and Glorieux, 2004). The clinical severity of OI is type II > type III > types IV = V = VI = VII > type I.
Table 91-1Expanded Sillence Classification of Osteogenesis Imperfecta ||Download (.pdf) Table 91-1 Expanded Sillence Classification of Osteogenesis Imperfecta
|Type ||Clincal Severity ||Typical Features ||Typically Associated Mutations |
|I ||Mild nondeforming osteogenesis imperfecta ||Normal height or mild short stature; blue sclerae; no dentinogenesis imperfecta ||Premature stop codon in COL1A1 |
|II ||Perinatal lethal ||Multiple rib and long bone fractures at birth; pronounced deformities; broad long bones; low density of skull bones on radiographs; dark sclerae ||Glycine substitutions in COL1A1 or COL1A2 |
|III ||Severely deforming ||Very short; triangular face; severe scoliosis; greyish sclerae; dentinogenesis imperfecta ||Glycine substitutions inCOL1A1 or COL1A2 |
|IV ||Moderately deforming ||Moderately short; mild-to-moderate scoliosis; greyish or white sclerae; dentinogenesis imperfecta ||Glycine substitutions in COL1A1 or COL1A2 |
|V ||Moderately deforming ||Mild-to-moderate short stature; dislocation of radial head; mineralized interosseous membrane; hyperplastic callus; white sclerae; no dentinogenesis imperfecta ||Unknown |
|VI ||Moderately to severely deforming ||Moderately short; scoliosis; accumulation of osteoid in bone tissue; fish-scale pattern of bone lamellation; white sclerae; no dentinogenesis imperfecta ||Unknown |
|VII ||Moderately deforming ||Mild short stature; short humeri and femora; coxa vara; white sclerae; no dentinogenesis imperfecta ||Unknown |
Type I is a form of dominantly inherited osteoporosis that leads to fractures. Affected patients have distinctly blue sclerae and between 35% and 50% have presenile conductive hearing loss or deafness. The earliest age of onset of hearing loss is 10 years, and 40% of affected adults eventually require hearing aids. Approximately one fifth of patients with type I osteogenesis imperfecta (OI) have kyphosis and scoliosis, although severe spinal curves are rarely seen. These patients also bruise easily. All patients with type I OI are able to walk independently. Type I is further subdivided into type IA, patients with normal teeth and type IB, patients who have dentinogenesis imperfecta. Only 10% of patients with type I OI have fractures that are identifiable at birth (Sillence, 1981). Patients affected with type I OI have a progressive loss of height due to platyspondyly and kyphosis. Birth weight and length are generally normal and short stature is of postnatal onset. These patients are also notable for a head size that appears large for height.
Patients with type II OI comprise the majority of cases detected pre- and post-natally. Type II has been further subdivided into types IIA, IIB, and IIC. In type IIA broad, crumpled femurs and continuous beading of the ribs are present. In addition, the patients are small for gestational age and have severe osteoporosis of the skull and face. In type IIB, there are minimal or no rib fractures present. Because the ribs are less severely affected, the chest configuration is more normal and the resulting respiratory distress is less severe. Type IIB is the only form with potential postnatal survival. In type IIC there are thin femurs and ribs with extensive fractures. In this form, the fetuses are very small for gestational age and severe osteopenia is present. However, many investigators state that distinction between the subgroups of type II is of limited value because all fetuses and infants with OI type II die during the perinatal period.
Type III is a rare form of OI, characterized by marked fragility and fractures of the long bones and skull, which are sometimes present at birth (Sillence et al., 1986). In utero, the defect in ossification of the skull is not as marked as it is in Type II. Posnatally, there are spine and long bone fractures, which result in progressive short stature and kyphoscoliosis. Although blue sclerae are present at birth, they fade with time. Hearing impairment is rare in type III OI. Affected patients have a triangle-shaped face with a wide bitemporal diameter. These patients are among the smallest of adults with OI. They have considerable difficulty walking. They suffer from multiple pulmonary complications.
Type IV is a dominantly inherited form of osteoporosis that leads to fractures. Variable deformity of the long bones exists, but affected patients have normal sclerae. There are no associated hearing abnormalites. Type IV has been further subdivided into IVA and IVB. In type IVA there is normal dentition and in IVB there is dentinogenesis imperfecta. To date, types V, VI and VII have not been associated with a specific prenatal presentation. All are associated with bone fragility, but none are known to be caused by mutations in COL1A1 or COL1A2. Patients with Type V OI have hypertrophic callus formation at fracture sites, calcification of the interosseous membranes between bones of the forearm, and a radio-opaque metaphyseal band adjacent to the growth plates. The distinctive feature of type VI OI is the histologic appearance of bone lamellae that resemble fish scales. Patients with type VI OI also accumulate excessive osteoid. Type VII OI is inherited in an autosomal recessive pattern, and it is characterized by proximal shortening of the humerus and femur (Roughley et al., 2003).
The incidence of type I OI is 1 in 28,500 livebirths, while type II occurs in 1 in 62,000 livebirths and type III occurs in 1 in 68,800 livebirths (Sillence et al., 1979). In Sillence et al.’s original article, they did not quote an incidence for type IV OI (Sillence et al., 1979). Rasmussen et al. (1996) identifed 16 cases of OI among 126,316 deliveries that occurred over a 15-year period in a single teaching hospital. These authors estimated a prevalence (with exclusion of high-risk patients) of 0.24 in 10,000 deliveries of type II OI and 0.4 in 10,000 of types II and III OI combined. OI has been described in all ethnic groups (Sykes et al., 1986).
The prenatal sonographic findings in OI are summarized in Table 91-2. The characteristic antenatal findings of OI include in utero fractures that occur with callus formation at the site of healing. These result in prenatally acquired long-bone deformities and significant limb shortening (Figure 91-1). Abnormalities of the fetal skull are the most striking findings in OI (Constantine et al., 1991). In addition, soft and fractured ribs contribute to a small thoracic circumference, which has been described as having a “champagne cork” appearance. The unusual clarity of intracranial structures is due to poor calvarial ossification (Figure 91-2). This has led to the term supervisualization (Andrews and Amparo, 1993). The compression of the fetal head by the ultrasound probe and the low echogenicity of the cranium should raise the suspicion of a skull dysplasia. However, this finding is not diagnostic for OI (Berge et al., 1995).
Table 91-2Prenatal Sonographic Findings in Osteogenesis Imperfecta (Listed in order of detection at earliest gestational age) ||Download (.pdf) Table 91-2 Prenatal Sonographic Findings in Osteogenesis Imperfecta (Listed in order of detection at earliest gestational age)
|Type ||Genetics ||Clinical Findings ||Ultrasound Findings ||First Ultrasound Detection |
|OI II lethal Perinatal ||Autosomal dominant ||Lethal perinatal type: Undermineralized skull, micromelic bones, “beaded” ribs on x-ray, bone deformity, platyspondyly ||Undermineralization, broad crumpled and shortened limbs, thin beaded ribs, fractures, angulation or bowing of long bones, normal appearing hands, deformable calvarium ||≥14 wks |
|OI III ||Autosomal dominant ||Progressively deforming type: Moderate deformity of limbs at birth, scleral hue varies, very short stature, dentinogenesis imperfecta (DI) ||Thin ribs, short limbs, fractures, undermineralized skull, long bone length falls away from normal at 16-18 weeks ||≥18 wks |
|OI IV ||Autosomal dominant ||Normal sclerae, mild/moderate limb deformity with fracture, variable short stature, DI, some hearing loss ||Rarely, long bone bowing and/or fracture ||After20wksbut not common |
|OI I ||Autosomal dominant ||Fractures with little or no limb deformity, blue sclerae, normal stature, hearing loss, DI ||Rarely, long bone bowing or fracture ||> 20 wks but not common |
|OI V ||Autosomal dominant ||Similar to OI IV plus calcification of interosseous membrane of forearm, radial head dislocation, hyperplastic callus formation ||Unknown ||Not described |
|OI VI ||Unknown ||More fractures than OI type IV, vertebral compression fractures, no DI ||Unknown ||Not described |
|OI VII ||Autosomal recessive ||Congenital fractures, blue sclerae, early deformity of legs, coxa vara, osteopenia ||Unknown ||Not described |
Sonographic image of an acutely angled humerus in a fetus with type II OI.
Cross-sectional view of a fetal head demonstrating significantly reduced skull ossification, resulting in unusually clear visualization of intracranial contents.
The following sonographic criteria have been proposed for type II OI: multiple fractures, demineralization of the calvarium, and a femoral length less than 3 SD below the mean for gestational age coupled with a wrinkled appearance of the long bones (Munoz et al., 1990). In a retrospective study of 459 fetuses and infants with bent femurs, 18.1% had OI (Alanay et al., 2007).
In the absence of a known family history of OI, most fetuses detected prenatally will have type II. The major diagnostic criteria for this type of OI include shortened deformed long bones, underossification of the cranial vault, which results in easily seen intracranial structures, an abnormal and varying skull shape, a small chest circumference with broad and irregular ribs (Figure 91-3), decreased fetal movements, and unusual fetal limb position (Constantine et al., 1991). In one case report, Morin et al. (1991) described a case of type IIA OI in one of dizygotic twins diagnosed at 27 weeks of gestation. In this report, the affected fetus was so translucent that only one twin could be seen on a plain radiograph. In another case report, D’Ottavio et al. (1993) described a case of type II OI in the fetus of a woman at 14 weeks of gestation who underwent routine transvaginal ultrasonography. In this fetus, both femurs were short and severely angulated because of fractures. Even at this early point in gestation, the fetal skull was noted to be hypoechogenic, and an abnormal curvature of the right radius was present.
Cross-sectional view of fetal thorax, demonstrating small chest circumference with undermineralized, beaded, crumpled ribs.
Prenatal diagnosis of OI types I and III is more difficult to make on a sonographic basis. Most of the cases described in the literature have been diagnosed in fetuses known to be at risk because of a positive family history. For example, Robinson et al. (1987) described a fetus at risk for type III OI that was followed with serial sonography. At 15 weeks of gestation, there was a low normal fetal femur length. By 20 and 22 weeks, however, shortening of the long bones and deformity of the femurs were noted. There was not an impressive decrease in ossification of the fetal skull. Several case reports of prenatal diagnosis of type I OI have appeared in families known to be at risk. Chervenak et al. (1982) described a fetus whose mother was affected with type I OI. This fetus had a normal sonographic examination at 20 weeks, but bowed femurs developed at 24 weeks. By 32 weeks of gestation, demineralization of the fetal skull was observed. No fractures were seen in utero, but a right femur fracture developed at 9 days of age, which was postulated to be due to the effect of intrauterine curvilinear stress on weakened bones.
Brons et al. (1988) reported on the sonographic diagnosis of OI in seven fetuses collected from the experience of three major teaching hospitals in the Netherlands. The gestational age at the time of scanning was between 15 and 34 weeks. The indications for sonography included large for gestational age (two fetuses), small for gestational age (two), previous child affected with OI (two), and routine anatomy scan (one). The biparietal diameter was normal in all seven. The abdominal circumference, however, was either normal or small for gestational age. In most cases, the chest circumference was narrow as compared with the abdominal circumference. The heart was noted to completely fill the chest. The limbs were the most severely shortened in type IIA OI. The prenatal diagnoses of types IIB, IIC, and III were made later in gestation than in type IIA (Brons et al., 1988).
The differential diagnosis includes other causes of severe skeletal dysplasia and demineralization. Conditions that most resemble OI on prenatal sonographic examination include hypophosphatasia (see Chapter 98) due to demineralization of the skull, narrow chest circumference, and shortened extremities. Fetuses with hypophosphatasia generally do not have in utero fractures, but they can have bowing or angulation of the long bones (Pauli et al., 1999). The condition that seems to be most commonly confused prenatally with OI is campomelic dysplasia (see Chapter 92). Sanders et al. (1994) described three cases of OI that mimicked campomelic dyplasia and one case of campomelic dysplasia that was prenatally diagnosed as OI. The overlap between campomelic dysplasia and OI relates to the severe bowing of the limbs (Alanay et al., 2007) (see Figure 91-1). Tibial bowing is more pronounced in campomelic dysplasia, but some cases of severe nonlethal OI have tibial bowing without obvious fractures. Conversely, acute angulation of the femur seen in campomelic dysplasia can suggest a fracture. Cranial deformities, such as bossing, hypertelorism, and hydrocephalus can exist in both syndromes. The presence of arm fractures with callus formation, cranial compressibility with unusually clear visualization of the intracranial contents, and asymmetry in the length of the limbs favors a diagnosis of OI. The presence of clubfeet, micrognathia, and hydronephrosis favors the diagnosis of campomelic dysplasia (Sanders et al., 1994). Another condition that can closely resemble OI is achondrogenesis (see Chapter 97).
ANTENATAL NATURAL HISTORY
Two major biochemical phenotypes of OI exist: patients with normal stature and blue sclerae (type I) secrete half of the normal amount of a normal type I procollagen and do not have the presence of identifiable abnormal molecules. Patients with short stature and fractures (types II, III, and IV) produce and secrete both normal and abnormal type I procollagen molecules (Wenstrup et al., 1990) (Table 91-1).
The perinatal lethal form, type II OI, is the result of heterozygous mutations of the COL1A1 and COL1A2 genes that encode the and the α2 (I) chains of type I collagen. These mutations act in a dominant negative manner because the mutant proα-chains are incorporated into type I procollagen molecules that also contain normal proa-chains. The abnormal molecules that are poorly secreted are more susceptible to degradation and impair formation of the extracellular matrix. The collagen fibers are abnormally organized, and mineralization is impaired (Cole and Dalgleish, 1995).
Histologic and biochemical abnormalities in the dermis obtained from babies with type II OI demonstrate fibroblasts with a dilated rough endoplasmic reticulum due to impaired secretion of mutant type I procollagen molecules. The collagen fibrils are smaller than normal. There is a reduced amount of type I collagen and a mixture of normal and mutant type I collagen. The bone is severely porotic. The normal cortical and trabecular bone is replaced by woven bone. There is an abundance of plump osteoblast surrounded by small amounts of extracellular matrix. The osteoblast may contain dilated rough endoplasmic reticulum. The growth plate is normal but the cartilage cores persist in the trabeculae. The bone matrix is decreased in type I collagen and increased in types III and IV collagen.
These histologic findings suggest that the skeletal dyplasia in type II OI results from the action of muscular forces on a skeleton weakened by a complex disorder of endochondral and intramembraneous ossification. The paucity of primary metaphyseal trabeculae and subperiosteal cortical bone leads to pathologic fracture of the immature fiber bone and abnormal attempts at fracture repair (Marion et al., 1993).
The pregnant patient carrying a fetus in which OI is suspected should be referred to a tertiary care center capable of sophisticated anatomic diagnosis. The patient should meet with a medical geneticist or genetic counselor, who will obtain a detailed family history and a three generation pedigree. Specific questions should be asked regarding the height of first-degree family members, the presence of deafness, the color of the sclerae in family members, and whether there is a history of fractures. If previously affected infants have been born to the family, pathologic studies (if available) should be reviewed. However, most cases of affected fetuses will occur in the setting of a negative family history. Consideration should be given to obtaining a prenatal radiograph if the bones are not adequately visualized on a level II sonogram. For a presumed diagnosis of OI, there is no indication to obtain chromosomal analysis. Chorionic villi and amniotic fluid cells can serve as equally good sources of DNA for mutation analysis of the COL1A1 and COL1A2 genes. Amniocytes, however, synthesize inadequate amounts of type I procollagen, so they are not useful for prenatal biochemical diagnosis (Byers et al., 2006). In contrast, chorionic villi do synthesize adequate amounts of procollagen for biochemical analysis.
In the setting of a positive family history and a fetal presentation of skeletal dysplasia at earlier than 20 weeks of gestation, the presumed diagnosis is types II or III OI. In such cases, the poor prognosis should be discussed with the parents, who can be offered the opportunity to electively terminate the pregnancy.
In patients with a known family history of OI, prenatal sonographic and biochemical diagnoses are available. CVS should be offered for biochemical studies in types II, III, and IV. Prenatal diagnosis for type I OI is only accurate if the underlying familial DNA mutation is known. Note that prenatal sonography in fetuses at risk for type I OI can be normal, even when the fetus is affected.
Pregnant women who themselves are affected with type I OI are at risk for uterine rupture during a spontaneous vaginal delivery due to decreased amounts of collagen (Carlson and Harlass, 1993). Previous reports exist of dysfunctional platelet aggregation and prolonged bleeding times in women affected with type I OI. In addition, these patients can have a hypermetabolic state with hyperthermia due to abnormal oxidative metabolism. These women are at increased risk for developing malignant hyperthermia with general anesthesia for cesarean delivery. For affected mothers and fetuses with nonlethal types of OI, the goal is to provide a safe and atraumatic delivery for both mother and fetus.
In a review of 167 pregnancies affected by a fetal diagnosis of OI, Cubert et al. (2001) showed that cesarean section delivery did not decrease fracture rates at birth in infants with nonlethal forms, nor did it prolong survival in those with lethal types. Interestingly, there was an unusually high rate of breech presentation at term (37% of cases). Because of the maternal morbidity and mortality associated with cesarean section delivery, these authors suggested that operative delivery be reserved for standard obstetric indications such as breech presentation or fetal distress.
There has been at least one published report of a 32-week female fetus with OI that underwent transplantation with allogeneic HLA-mismatched male fetal mesenchymal stem cells (MSCs) (Le Blanc et al., 2005). In total, 6.5 × 106 fetal MSCs were injected into the umbilical vein and there were no post-procedural complications. Spontaneous rupture of the membranes occurred at 35 weeks. Delivery was by cesarean section. Postnatal radiography indicated multiple healed fractures and a new right femoral fracture, which was stabilized prior to discharge. Treatment with pamidronate was initiated at 4 months. At 9 months a bone biopsy showed regularly arranged bone trabeculae lined by a columnar layer of normal osteoblasts. FISH analysis was performed using centromeric X- and Y-chomosome specific probes. This showed that 0.3% of cells staining positive for osteopontin or osteocalcin were donor derived. At 2 years of age the patient was small, but had normal development. She had three fractures, but did not require any overnight hospitalization for treatment. This report was the first demonstration of donor MSC engraftment into an immunocompetent human fetus (Le Blanc et al., 2005).
The majority of newborns who present with fractures have type II OI. Postnatal radiography may aid in the determination of a definite diagnosis (Figure 91-4). The prognosis is poor for these infants, and the cause of death is usually secondary to respiratory difficulties associated with mechanical failure of the chest wall (Carlson and Harlass, 1993). Supportive care is indicated for infants likely to have type II OI. A skin biopsy should be obtained for definitive biochemical and/or DNA analysis. A fibroblast culture can be sent to a referral laboratory for analysis of quantity and quality of type I collagen chains.
Postnatal radiographs of a newborn infant with type II OI. (Left) Multiple fractures are present in the long bones. The ribs are poorly mineralized. (Right) Close-up skull radiograph in same infant, demonstrating the lack of ossification that is characteristically observed in prenatal sonographic studies.
For infants with nonlethal forms of OI, the mainstays of therapy are physical therapy, rehabilitation, and orthopedic surgery. These infants are likely to benefit from specific exercise, bracing, and ambulation training (Binder et al., 1993). Only routine hearing screening is recommended, as the hearing abnormalities characteristic of type I OI do not develop until later childhood or early adulthood. Parents of affected newborns should be counseled that blue sclerae represent a change in physical appearance and do not affect visual acuity.
Infants who survive the newborn period may benefit from bisphosphonate treatment (Glorieux, 2007). Bisphosphonates are pyrophosphate analogs that inhibit osteoclast-mediated bone resorption. In response to bisphosphonate therapy, bone mass increases due to a relative increase in osteoblastic activity. The first large scale clinical trial of children with severe OI who were treated with intravenous pamidronate (1 mg/kg/d) for 3 consecutive days every 4-6 months was published in 1998 (Glorieux et al., 1998). More than 50% of the study population showed improved mobility, fracture incidence was reduced to 1.7 per year, and the mean annual increment in spinal bone density was 42%. These results were quickly validated by others (Astrom and Soderhall, 1998). Many studies have now shown that treatment increases bone mineral density by increasing cortical bone width and the number of trabeculae present (Roughley et al., 2003). Most important to the affected individuals is the fact that chronic bone pain decreases with treatment. This permits increased mobility and improvement in muscle strenth. Several hundred patients have received bisphosphonate therapy for periods of up to 7 years and it is now considered to be standard care for children with moderate to severe OI (Glorieux, 2007). No benefits have been established for mild cases in which negative effects, such as decrease in bone remodeling rate, reduction in growth plate cartilage resorption, and delay in osteotomy site healing may outweigh any positive effects. Bisphosphonates also stay in bone for many years following treatment. Although there is potential concern about teratogenic effects in fetuses of treated women with OI, at least one report describing two babies of two pregnant women with types I and IV OI did not show any skeletal modeling abnormalites consistent with in utero pamidronate exposure (Munns et al., 2004). Still unanswered questions regarding bisphosphonate treatment include criteria used to start treatment, optimal length of treatment, duration of treatment effect, dose and frequency of administration, and criteria for discontinuing treatment (Shaw and Bishop, 2005).
It is important to realize that bisphosphonate therapy does not cure OI. Alternative approaches include cellular therapy, such as bone marrow (Horwitz et al., 2002) or mesenchymal stem cell transplantation (Le Blanc et al., 2005), as well as ex vivo gene therapy (Chamberlain et al., 2004).
The orthopedic manifestations of the nonlethal forms of OI consist of bone fragility, with resulting fractures, deformity, and scoliosis. When these fractures occcur, the bones heal with callus formation and a successive slight malalignment. Repetitive fractures of the same bone may lead to a significant deformity. Eventually, decreased range of motion of the joints and contractures can occur. Fractures are treated with immobilization or intramedullary rodding, which consists of a surgical insertion of a pin into the long bones to stabilize them. Scoliosis occurs in 30% to 70% of patients with OI. This requires body bracing with jackets that prevent progression of the curve.
Stabilization of the long bones in affected patients may help gross motor devlopment (Engelbert et al., 1995). In one study, the gross motor development of 10 children with type III OI was retrospectively studied. All patients were noted to have a severe delay in gross motor development. Interestingly, the sequence of achieving developmental milestones was different as compared with normal individuals. The static milestones, such as complete head control and sitting without support, developed earlier in patients with OI than the dynamic milestones, such as lifting the head in the prone position and rolling over from supine to prone. These authors hypothesize that affected infants protect themselves from developing fractures. They compared the gross motor development in their patients as a function of when they received intramedullary rodding. They found that the patients who underwent intramedullary rodding of the lower extremities before the age of 3.5 years had better neuromotor development. They recommended that before being considered for surgery, affected individuals should be able to sit unsupported. In this series, the median age of achievement of this milestone was 2.1 years. The recommendation, therefore, was that intramedullary rodding should be performed between 2 and 3.5 years of age to permit improved long-term ambulation (Engelbert et al., 1995).
Type I OI is the only type in which blue sclerae remain present throughout life. All types of OI put the patient at risk for fractures and scoliosis. The scoliosis may result in decreased cardiopulmonary reserve and predisposition to recurrent respiratory infections. Patients who have dentinogenesis imperfecta have abnormal dentin. The enamel is unable to adhere adequately and chips and erodes easily. Although the teeth are abnormal in appearance, the incidence ofdental caries is not increased (Varni and Jaffe, 1984). Presenile hearing loss may develop in early adulthood in patients with type I OI. This is a conductive hearing loss due to the stapes being partially replaced by fibrous tissue.
Patients affected with OI also have an increased incidence of inguinal and umbilical hernias. Hypercalciuria is a common finding that correlates with the severity of the skull disease; renal function and postnatal renal sonographic studies are otherwise normal (Chines et al., 1995).
Paterson et al. (1996) reviewed the life expectancy for 743 patients with OI. Patients with type II, the perinatal lethal form, were excluded from the study. Their patient population included 383 patients with type IA, 77 with type IB, 123 with type III, 90 with type IVA, and 70 with type IVB. Patients with type IA OI had no difference in mortality from the general population. Types IB, IVA, and IVB had a modestly reduced life expectancy as compared with the normal population. Only patients with type III had significant impairment in their life expectancy. Of the 26 deaths reviewed in this study, 19 occurred before the age of 10 years. In a follow-up study, McAllion and Paterson (1996) reviewed 79 causes of deaths in patients affected with OI. Of the 79, 46 were due to respiratory causes and these occurred mainly in patients affected with type III. The respiratory deaths were primarily due to infection, and cardiac failure resulted from kyphoscoliosis. Other causes of death that were directly or indirectly due to OI included basilar invagination of the skull and lethal intracranial bleeding. These authors stress the importance of obtaining prompt care for respiratory infections and prevention of head trauma in affected patients.
GENETICS AND RECURRENCE RISK
In about 90% of cases of OI there is a mutation in either COL1A1 or COL1A2 (Byers et al., 2006). The loci COL1A1 (α1-chain) and COL1A2 (α2-chain) have been mapped to human chromosomes 17 and 7, respectively (Sykes et al., 1986).
The severity of the clinical phenotype is related to the type of mutation, its location on the α-chain, the surrounding amino acid sequences, and the level of expression of the mutant allele (Cole and Dalgleish, 1995). Point mutations resulting in a substitution of Gly residues in Gly-X-Y amino acid triplets of the triple helical domain of the α1 or (I) α2 (I) chains are the most frequent mutations. They interrupt the repetitive Gly-X-Y structure that is mandatory for formation of a stable triple helix.
Of affected individuals studied at the molecular level, it has been shown that most babies have their own private de novo mutation in type II OI. Few deletions have been identified. Exon skipping mutations are more common, which usually maintain the translational reading frame and the repetitive Gly-X-Y amino acid triplet structure. Deletions, skipping mutations, and insertions in triple helical domains produce abnormal alignment of the proα-chains. This interrupts the mandatory Gly-X-Y structure and may produce a bulge or a kink in the molecule (Cole and Dalgleish, 1995).
In type I OI, haploinsufficiency of COL1A1 results in a 50% reduction of type I collagen. Most patients with type I have decreased mRNA levels of COL1A1 (Willing et al., 1993). In types II, III, and IV there are dominant negative mutations that synthesize abnormal procollagen chains that bind to normal chains, thereby destroying their biologic activity. The dominant negative mutations are more harmful than the null mutations characteristic of type I OI. In fact, the goal of gene therapy is to degrade the mutant mRNA or disrupt the mutant gene to create essentially a null mutation. This would convert the more severe phenotypes seen in types II, III, and IV to a type I phenotype.
Most mutations associated with OI are dominantly inherited. Autosomal recessive phenotypes are rare, except in consanguinous families, and in type VII OI. Even with a negative family history, type II OI that results from an apparently new dominant mutation has an empiric recurrence risk of 7% (Byers et al., 1988). This risk increases significantly for unaffected parents after the diagnosis of a second affected fetus, because it identifies them as mosaic carriers of a mutant allele. Germline mosaicism for a mutation in one of the genes that encodes a chain of type I procollagen is surprisingly common. Both germline mosaicism and somatic mosaicism can occur (Sykes, 1990). In one report, a mutant allele was present in 80% of lymphocytes and 100% of skin fibroblasts of a parent of an affected child. Thus, the mutation must occur very early in embryonic development, before the somatic cell lines and germ cell lines segregate. In a prospective ascertainment of the recurrence risk for couples with one previous affected fetus for whom prenatal diagnostic studies were undertaken, only 1 in 50 pregnancies (2%) was affected (Pepin et al., 1997) (Table 91-3).
Table 91-3Prospectively Ascertained Recurrence Risk in 129 Prenatal Diagnoses of OI ||Download (.pdf) Table 91-3 Prospectively Ascertained Recurrence Risk in 129 Prenatal Diagnoses of OI
| || No. of Cases || Affected || Risk |
|Type I |
|Affected parent ||16 ||11 ||68% |
|Type II |
|One prior affected fetus ||50 ||1 ||2% |
|Two prior affected fetuses ||7 ||2 ||28% |
|Type III or IV |
|Affected parent ||22 ||11 ||50% |
|One prior affected fetus ||31 ||1 ||3% |
|Two prior affected fetuses ||3 ||0 ||–- |
Prenatal diagnosis in subsequent pregnancies can be performed by sonography, biochemical, or molecular analysis. In type II OI, transvaginal sonography can readily detect abnormalities during the late first or early second trimester. If the molecular defect is unknown in a specific family, biochemical analysis of the collagen and procollagen molecules in fibroblasts obtained from chorionic villi sampling can provide first trimester prenatal diagnosis in types II, III, and IV OI (Pepin et al., 1997). Specific evidence can be sought of post-translational overmodification of the proα1(I)-chains of type I procollagen. Cells from chorionic villi produce type I collagen chains with the same electrophoretic abnormalities as skin collagen (Grange et al., 1990). Chorionic villus sampling is the prenatal diagnosis method of choice for families with a positive history of types II, III, and IV OI. In pregnancies at risk for type I OI, identification of reduced amounts of type I collagen is inaccurate. Prenatal diagnosis of type I OI is best achieved by direct mutational analysis (Pepin et al., 1997).
The accuracy of prenatal diagnosis of OI is largely dependent on the prior study of an affected individual in a specific family (Pepin et al., 1997). Diagnostic information is usually available within 20 to 30 days of chorionic villus sampling using biochemical techniques and within 10 to 14 days when molecular analysis is utilized (Pepin et al., 1997).
RS. Angulated femurs and the skeletal dysplasias: experience of the International Skeletal Dysplasia Registry (1988-2006). Am J Med Genet A. 2007;143A:1159-–1168.CrossRef
EG. In utero clue to congenital osteogenesis imperfecta. AJR Am J Roentgenol. 1993;60:212.CrossRef
S. Beneficial effect of biphosphonate during five years of treatment of severe osteogenesis imperfecta. Acta Paediatr. 1998;87:64-–68.
P. Prenatal diagnosis of osteogenesis imperfecta. Acta Obstet Gynecol Scand. 1995;74:321-–323.
LH. Rehabilitation approaches to children with osteogenesis imperfecta: a ten year experience. Arch Phys Med Rehabil. 1993;74:386-–390.
JT, Van Der Harten
et al.. Prenatal ultrasonographic diagnosis of osteogenesis imperfecta. Am J Obstet Gynecol. 1988;159:176-–181.
M. American College of Medical Genetics. Genetic evaluation of suspected osteogenesis imperfecta (OI). Genet Med. 2006;8:383-–388.
RC. Perinatal lethal osteogenesis imperfecta (OI type II): a biochemically heterogeneous disorder usually due to new mutations in the genes for type I collagen
. Am J Hum Genet. 1988;42:237-–248.
FE. Management of osteogenesis imperfecta in pregnancy: a case report. J Reprod Med. 1993;38:228-–232.
et al.. Gene targeting in stem cells from individuals with osteogenesis imperfecta. Science. 2004; 303:1198-–1201.
et al.. Antenatal sonographic findings of osteogenesis imperfecta. Am J Obstet Gynecol. 1982; 143:228-–230.
M. Hypercalciuria in osteogenesis imperfecta; a follow-up study to assess renal effects. Bone. 1995;16:333-–339.
A. Prenatal diagnosis of severe osteogenesis imperfecta. Prenat Diagn. 1991;11:103-–110.
PH. Osteogenesis imperfecta: mode of delivery and neonatal outcome. Obstet Gynecol. 2001;97:66-–69.
G. Early prenatal ultrasonographic diagnosis of osteogenesis imperfecta: a case report. Am J Obstet Gynecol. 1993;169:384-–385.
RHJM. Intramedullary rodding in type III osteogenesis imperfecta. Acta Orthop Scand. 1995;66:361-–364.
FH. Experience with bisphosphonates in osteogenesis imperfecta. Pediatrics. 2007;119:S163-–S165.
et al.. Cyclic administration of pamidronate
in children with severe osteogenesis imperfecta. N Engl J Med. 1998;339:947-–952.
JC. Analysis of cultured chorionic villi in a case of osteogenesis imperfecta type II: implications for prenatal diagnosis. Am J Med Genet. 1990;36:258-–264.CrossRef
et al.. Isolated allogeneic bone marrow-derived mesenchymal cells engraft and stimulate growth in children with osteogenesis imperfecta: Implications for cell therapy of bone. Proc Natl Acad Sci USA. 2002;99:8932-–8937.
et al.. Fetal mesenchymal stemcell engraftment in bone after in utero
transplantation in a patient with severe osteogenesis imperfecta. Transplantation. 2005;79:1607-–1614.
Fs. Skeletal dysplasia in perinatal lethal osteogenesis imperfecta: a complex disorder of endochondral and intramembranous ossification. Clin Orthop Relat Res. 1993;293:327-–337.
CR. Causes of death in osteogenesis imperfecta. J Clin Pathol. 1996;49:627-–630.
V. Prenatal diagnosis of lethal osteogenesis imperfecta in twin pregnancy. Clin Genet. 1991;39:467-–470.
Fh. Maternal and fetal outcome after long-term pamidronate
treatment before conception: a report of two cases. J Bone Miner Res. 2004;19:1742-–1745.
MS. Osteogenesis imperfecta type II: prenatal sonographic diagnosis. Radiology. 1990;174:181-–185.
MP. Mild hypophosphatasia mimicking severe osteogenesis imperfecta in utero: bent but not broken. Am J Med Genet. 1999;86:434-–438.CrossRef
Ph. Strategies and outcomes of prenatal diagnosis for osteogenesis imperfecta: a review of biochemical and molecular studies completed in 129 pregnancies. Prenat Diagn. 1997;17:559-–570.
Lb. Epidemiology of osteochondrodysplasias: changing trends due to advances in prenatal diagnosis. Am J Med Genet. 1996; 61:49-–58.CrossRef
Dl. Prenatal diagnosis of osteogenesis imperfecta type II. Prenat Diagn. 1987;7:7-–15.
FH. Osteogenesis imperfecta-clinical and molecular diversity. Eur Cell Mater. 2003;5:41-–47.
S. Osteogenesis imperfecta and campomelic dysplasia: difficulties in prenatal diagnosis. J Ultrasound Med. 1994;13:691-–700.
DO. Osteogenesis imperfecta: an expanding panorama of variants. Clin Orthop Relat Res. 1981;159:11-–25.
Dl. Osteogenesis imperfecta type III: delineation of the phenotype with reference to genetic heterogeneity. Am J Med Genet. 1986;23:821-–832.CrossRef
DM. Genetic heterogeneity in osteogenesis imperfecta. J Med Genet. 1979;6:101-–116.CrossRef
N. Osteogenesis imperfecta is linked to both type I collagen
structural genes. Lancet. 1986;2:69-–72.
M. Osteogenesis imperfecta: the basics. Pediatr Nurs. 1984;10:29-–33.
Ph. Distinct biochemical phenotypes predict clinical severity in nonlethal variants of osteogenesis imperfecta. Am J Hum Genet. 1990;46:975-–982.
PH. Molecular heterogeneity in osteogenesis imperfecta type I. Am J Med Genet. 1993;45:223-–227