Журнал «Боль. Суставы. Позвоночник» 3 (07) 2012

Osteogenesis imperfecta (OI), also known as brittle bone disease, is a genetic disorder of connective tissue characterized by fragile bones and a susceptibility to fracture from mild trauma. The clinical range of this condition is extremely broad, ranging from lethal cases in the perinatal period to cases that may be difficult to detect in later life and can pretend early osteoporosis. Patients with OI may have growth deficiency, wormian bones, scoliosis and extraskelatal manifestations such as defective tooth formation (dentinogenesis imperfecta), hearing loss, macrocephaly, blue sclerae, barrel chest up to hyperlaxity of joints and ligaments [1].

Аbout 85–90 % of patients with clinical OI have abnormalities of type I collagen, the major structural component of the extracellular matrix of bone, skin and tendon, caused by an autosomal dominant pattern. Classical OI has been described by Sillence in 1979 [2]. He divided OI into four subtypes, based on clinical and radiographic features. Consecutively molecularbiological studies have shown that the mild Sillence type I OI is caused by quantitiative defects in type I collagen [3]. Individuals with type I OI synthesize a reduced amount of structurally normal type I collagen because of a null COL1A1 allele, with a relative increase in the COL2/COL1 ratio [3].

The moderate and severe types of autosomal dominant OI are caused by structural defects in one of the two chains encoding for the type I collagen heterotrimer [4]. In types II, III and IV a mixture of normal collagen and collagen with a structural defect is synthesized.

Developments in recent years have shown the genetic background of relatively rare recessive OI [5]. Autosomal recessive OI is caused by defects in proteins of the prolyl­3­hydroxylation (P3H1) complex, CRTAP [6, 7], LEPRE1 and PPIB or helical folding (FKBP10, SERPINH1) [4]. About 5 % of OI cases are not caused by defects in either, type I collagen or the P3H1 complex.

Clinical Characteristics and Types of OI

Because the types of OI vary widely in symptoms, clinical appearance and in their onset, the diagnosis varies with the age of the individuals. A positive family history is usually not present, because appearance of classical OI in children of unaffected parents is caused by parental mosaicism [8].

Prenatally, severe types may be difficult to distinguish from thanatophoric dysplasia (missense mutations in fibroblast growth factor receptor­3), campomelic dysplasia (mutations in or near the SOX9 gene), and achondrogenesis type I (type 1A etiology is unknown, type 1B is caused by mutations in the SLC26A2 gene). The key diagnostic element for OI is the generalized nature of the connective tissue defect with facial features (flat midface, triangular shape, blue sclerae, yellowish or opalescent teeth), relative macrocephaly, thora­cic deformations, like barrel chest or pectus excavatum, joint laxity, morphologic changes of the vertebrae and growth deficiency, present in variable combinations in each case. When a diagnosis is still in doubt, collagen biochemical tests and DNA sequencing provide helpful information on the presence of a mutation.

The classification proposed by Sillence [2] is based on clinical and radiographic criteria that distinguished four types. The Sillence types have autosomal dominant inheritance. Although, laboratory practice has subsequently developed, the classification is still in use. More recently the types of OI have been extended to the autosomal recessive types V through XI [4] (table 1), although they are defined by different criteria than types I­IV (OMIM 166200, 166210, 259420, 166220). Clinical, the recessive forms of OI overlap with types I and III OI but have the distinction of white sclerae [6, 9].

OI Type I is the mildest form of the disorder. Fracture occurrence starts postnatally and usually decreases or even stops after puberty. Fracture risk in OI increases again after menopause in women and over 60 years of age in men [10].

Individuals with type I OI have blue sclerae and often bruise easily. They may have hearing loss or joint hyperextensibility. Growth reduction and deformities of the long bones are mild. Based on the presence of dentinogenesis imperfecta, type I has been divided into subtypes A and B.

OI Type II shows a very high perinatal lethality, however, survival to one year or more has been noted. These individuals are often born prematurely and are small for gestational age. Legs are usually held in the frog leg position with hips abducted and knees flexed. In X­rays long bones are extremely osteoporotic, with in­utero fractures and abnormal modeling. The skull is severely undermineralized with wide­open fontanels. The sclerae are blue­gray. Bones are composed predominantly of woven bone without haversian canals or organized lamellae.

OI Type III is known as the progressive deforming type. Most individuals with type III OI survive childhood with severe bone dysplasia. The presentation at birth may be similar to mild type II OI spectrum. They have extremely fragile bones and sustain up to hundreds of fractures over a lifetime. The long bones are easily deformed from normal muscle tension, and subsequently fractures occur. These individuals have extreme growth deficiency. Almost all type III cases develop scoliosis. Radiographically, bone abnormalities such as a flaring of the metaphy­ses and «popcorn» formation at growth plates are seen in addition to osteoporosis. The individuals require intensive physical rehabilitation and orthopedic care. Many of them are bounded to wheelchairs.

OI Typ IV is the moderately severe Sillence form. The diagnosis may be made at birth or delayed until school ages. Scleral hue is variable. These children often have several fractures a year and bowing of their long bones. Again, fractures decrease or even stop after puberty. Essentially, all type IV individuals have short final stature, which is often in the range of pubertal children. Many of these children are responsive to growth hormone by significant additional height. X­rays of the bones show osteoporosis and mild modeling abnormalities. Many deve­lop vertebral compressions and scoliosis. With consistent rehabilitation and orthopedic management, these individuals should be able to attain independent mobility (fig. 1).

For Type V–XI the Sillence numeration has been continued, but they are based on different criteria than the Sillence types. Clinically they present a phenotype, according to Sillence type IV. Type V (OMIM %610967) and type VI (OMIM %510968) are defined by histologic and cli­nical/radiographic signs and have an unknown etiology.

OI Typ V OI is associated with radiographical dense band adjacent to the growth plate of long bones. Patients with OI V develop hypertrophic callus formation at the sites of fractures or surgical procedures. They show calcification of the membrane between the radius and ulna, leading to restric­ted rotation. Patients have normal teeth and white sclerae. [11]

OI Type VI can be distinguished only on tissue level on bone biopsies [12]. The lamellae show «fish­scale»­like appearance under the microscope. OI VI is characterized by an increased amount of non­mineralized osteoid with a relatively late onset of fractures. Teeth and sclera development is normal, the skeletal disease seems to be moderate to severe [12]. OI VI show only a poor response to bisphosphonate therapy [13].

OI Type VII is caused by defects in cartilage­associated protein (CRTAP) [6, 7, 14]. These individuals have rhizomelia and a moderate bone disease, associated with a hypomorphic mutation in CRTAP. Null mutations in CRTAP have been shown to cause a lethal form of OI, with white sclerae, rhizomelia and a small to normal cranium.

OI Type VIII is caused by defects in P3H1 (encoded by LEPRE1). There is considerable overlap in the phenotypes of OI VII and VIII. Null mutations in LEPRE1 result in a phenotype that overlaps types II and III OI, but has distinct features like white slerae, extreme growth deficiency, and undermineralization. [9]

OI Type IX is caused by a gene defect in PPIB (CyPB). The phenotype is moderate to severe [4].

OI Type X is associated with SERPINH1 missense mutations (collagen chaperone HSP47). The only child reported with HSP47 deficiency had a severe OI phenotype, including blue sclerae, dentinogenosis imperfecta, transient skin bullae, pyloric stenosis and renal stones [4].

OI Type XI is caused by FKBP10 frameshift mutations. Patients have deforming OI including long bone fractures, ligamentous laxity, platyspondyly and scoliosis. Sclerae and teeth are normal [4].

Recently also the Bruck Syndrome type 2, Caffey Disease and Osteoblast Maturation Defects have been added as unclassified OI­like or collagen­based disorders to the OI classification system.

Radiologic Findings

Long bones have thin cortical bone and a gra­cile appearance. In moderately to severely affec­ted patients, long bones have bowing and modeling deformities, including cylindrical configuration from an apparent lack of modeling, metaphyseal flaring, and “popcorn” appearance at the metaphyses [15]. Long bones of the upper extremity are often seen with milder deformities than those of the lower extremity, even without weight bea­ring. Vertebrae often have central compressions even in mild type I OI. These often appear first at the T12­L1 level, consistent with weightbearing stress. In moderate to severe OI, vertebrae show central and anterior compressions and may appear compressed throughout. The compressions are generally consistent with the patient´s L1­L4 DXA Z­score but do not correlate in a straightforward manner with scoliosis. In the lateral view of the spine it is difficult to assess the asymmetry of vertebral collapse, which, along with paraspinal ligamentous laxity, is generally the cause of OI scoliosis. The skull of OI patients, with a wide range of severity, has wormian bones, although this is not unique to OI. Patients with type III and IV OI may also have platybasia, which should be followed with periodic CT studies for basilar impression and invagination [16].

The skeletal radiographic appearances of only a few infants and children with OI types VII and VIII have been described [6, 7, 9]. Both groups have extreme osteoporosis and abnormal long bone modeling, lea­ding to a cylindrical appearance. The bone material appears cystic and disorganized. In surviving children with type VIII there is a flaring of the metaphyses.

Laboratory and Histomorphometric Findings

Parameters of bone metabolism are generally in the normal range. However, bone turnover has also been described to be high [17] or even low­normal in OI [18]. Alkaline phosphatase may be elevated after a fracture and is lightly elevated in type VI [12]. ­Acid phosphatase is elevated in type VIII OI and can logically be expected to be elevated in type VII. Hormones of the growth axis usually have normal levels [19]. P1NP levels are decreased due to pathophysio­logy of disease [20]. Lower levels were found for OI I in comparison to OI III or OI IV due to the reduced amount of collagen [21]. Serum­CTX levels are usually in normal range [17] or decreased [20].

Bone histomorphometry shows defects in bone modeling and in production and thickening of trabeculae [22]. Cortical width and trabecular bone volume are decreased in all types. Trabecular number and width are decreased. Bone remodeling is increased, as are osteoblast’s and osteoclast’s surfaces [23]. Under polarized light, the lamellae of OI bone are thinner and less smooth than in controls. Mineral apposition rate is normal, crystal disorganization may constribute to bone weakness.

BMD Measurements

BMD measurements by DXA (L1­L4) are useful over a wide age and severity range of OI [24]. It aids diagnosis in milder cases and facilitates longitudinal follow­up in moderate to severe forms. There is a general correlation of Z­score and severity of OI. Type I individuals are generally in the ­1 to ­2 range, type IV Z­score cluster in the ­2 to ­4 range, whereas type III spans ­3 to ­6. Children with type VII OI have ­6 to ­7 Z­score. It is important to remember that the Z­score compares the mineral content of the bone being studied to bone with a normal matrix structure and crystal alignment. In OI, many mutations result in irregular crystal alignment on the abnormal matrix, in addition to reduced mine­ral quantity.

Nevertheless, also cases of high BMD­OI with elevated DXA values in comparison to controls and other patients with osteogenesis imperfecta due to increased bone mineralization were reported [25]. These cases were also associated with increased bone fragility. An increase in mineralization density of the bone matrix is a valuable explanation why the incidence of osteoporosis is lower than expected in adult OI.

Dual Energy X­ray absorptiometry (DXA) is currently the standard method for measuring bone density for the diagnosis and follow up of the most bone diseases. However, DXA does not measure bone quality, which includes bone geometry, histomorphometry and mechanical properties. Ano­ther limitation is, that DXA is unable to reliably differentiate between cortical and trabecular bone. In addition, DXA measurements after surgical intervention (e.g. kyphoplasty, surgical stabilisation of the vertebra) are unfeasible and degenerative changes and vertebral fractures complicate the analysis of DXA (see figure 2). In these cases the evaluation of the BMD in the hip region, domina­ted by cortical bone, is the only possible measuring site. Hip structure analysis (HSA) as an estimate of bone strength including cross­sectional moment of inertia (CSMI), cross­sectional area (CSA) and femoral strength index (FSI) is known to correlate with bone mass distribution and fracture occurrence. We could recently show the advantages of the combination of BMD and hip structure ana­lysis by DXA in adult patients with OI [26]. Especially CSA and CSMI could be an additional helpful means in estimating bone strength at the femoral neck in OI.

A relatively new technique for the assessment of volumetric BMD and bone structure measurement is the high resolution peripheral computed tomography (HR­pQCT). HR­pQCT seems to be a promi­sing tool for non­invasive structure evaluation in adult patients with OI. Performed at the radius and the ti­bia in patients with mild OI type I, HR­pQCT detected reduced areal BMD, volumetric BMD, bone area and trabecular number, compared to healthy controls [27]. Alterations of bone structure, measured by HR­pQCT, in moderate and severe OI types are under research at present (fig. 3).

Treatment Options

It has to be noted, that there is currently no cure for OI. Targets of treatment include an increase in BMD, a decrease in fracture rate, a reduction of pain, improvement of mobility as well as an increase in growth velocity.

Medical Therapy

Growth hormone (GH) therapy in patients with OI suggests an acceleration of short­term height velocity. The severe growth deficiency of OI is responsive to exogenous growth hormone administration in about one half of cases of type IV OI [28] and most type I OI [29]. Some trea­ted children can attain heights within the normal range. In addition, an increase in BMD, bone vo­lume per total volume (BV/TV) and bone formation rate (BFR) were shown. However, GH increases bone turnover, which seems to be counterproductive in a high bone turnover state like OI. Additionally, a non­responder status has been described [30].

Numerous controlled trials have shown the benefits and limitations of bisphosphonate treatment for OI [31–33]. The trabecular bone of vertebral bodies has the most positive response. BMD is increased, although the functional meaning of this measurement is difficult to assess because it also includes retained mineralized cartilage. More importantly, the vertebral ability to resist compressive forces is shown as increased vertebral area and decreased central vertebral compressions. The effect of bisphosphonate treatment on predominantly cortical long bone is more equivocal. There is a combination of increased stiffness and load bearing that is balanced by weakened bone quality [34].

Primary, pamidronate was administered in patients with OI [35]. An increase in vertebral height and a fracture reduction were noted in studies with cyclic pamidronate in children with severe OI [35]. However, also zoledronate acid showed satisfactory results regarding BMD and fracture risk reduction [36, 37].

Oral bisphosphonates like risedronate increase BMD at the lumbar spine, but not at the hip. However, fracture incidence and bone pain are not improved with oral bisphosphonates [38]. In addition, oral bisphosphonates do not seem to be appropriate in children, due the gastro­oesophageal side effects. In summary, it has been shown that there is a trend toward reduced fracture incidence or a reduced relative risk rather than a clear statistical benefit. The functional changes in ambulation, muscle strength and bone pain reported in uncontrolled trails have been shown to be similar to placebo. The prolonged half­life and recirculation of pamidronate in children up to 8 years after treatment cessation may pose paediatric specific skeletal and reproductive risks [39]. Prolonged or high­dose administration to children can induce defective bone remodeling [40] and may lead to accumulation of bone microdamage. Delayed osteotomy healing was noted at conventional doses [41]. The current management of bisphosphonates for OI is to treat for 2–3 years and then reduce the dose or discontinue the drug but continue to follow the patient.

Future prospects in OI treatment

The RANKL antibody denosumab is an established therapy in postmenopausal osteoporosis [42] and bone metastases [43]. Recently, denosumab has been determined in a small cohort of rare OI VI, who show poor response to bisphosphonate treatment [13]. Denosumab was injected subcutaneously every three months and caused a decrease of bone resorption, greater than the previous bisphosphonate the­rapy. Bisphosphonates stay in the skeleton for many years. One of the advantage of denosumab could be faster elimination [13].

However, further studies have to focus on effects of a long­term treatment and fracture risk prediction in bigger cohorts and different types of OI.

Osteoanabolic therapy with teriparatide (TPTD) is known to increase BMD and reduce vertebral fracture risk in male and female patients with osteoporosis [44]. In 2009 a phase four, multicentre, placebo­controlled study was enrolled to determine the effectiveness of TPTD in adult patients with OI (unpublished data). We could recently show the benefit of short­term TPTD treatment in fracture healing in a male patient with OI type IV and a five­fold pelvic fracture (ahead of print).

The Glycoprotein sclerostin is a potent inhibitor of bone formation. A sclerostin antibody was lately developed for the treatment of osteoporosis, showing promising data for this anabo­lic treatment in preclinical studies [45]. Currently the sclerostin antibody was evaluated in the Brtl/+ mouse knock­in model for moderately severe Type IV OI with a point mutation on COL1A1 [46]. Two weeks of treatment increased trabecular bone mass, resulted from increased trabecular thickness, but not number. Additionally, sclerostin antibody treatment significantly increased ultimate load and stiffness. Functional outcomes including ultimate load and stiffness were improved to levels not significantly diffe­rent from wild type mice [46]. However, the role of sclerostin in patients with OI remains unclear. Further investigations are needed on this field to confirm these preliminary data.

Mesenchymal stem cells (MSC) transplantation could possibly correct genetic disorders like OI [47]. Horwitz et al. performed bone marrow transplantations in three children with severe OI. Specimen of iliac bone biopsies, before MSC­transplantation, showed the characteristic appearance of high bone turnover, disorganized osteocytes, enlarged lacunae and a decreased number of osteoblasts. Seven months after MSC­transplantation a reduced number of osteocytes, linearly organized osteoblasts, improved bone formation and mineralisation were found. Additionally, bone mineral content, measured by DXA, increased by 28 g (median) during the first 100 days after transplantation [47]. Taking into account the small sample size of this study, MSC­transplantation could be a robust treatment in patients with OI. However, adverse effects of MSC such as tumour modulation, malignant transformation and immunosuppressive property have to be considered [48].

Physical Therapy and Orthopedic Surgery

Early and consistent rehabilitation intervention is the basis for maximizing the physical potential of individuals with OI [49, 50]. Physical therapy should begin in infancy for the severest types, promoting muscle strengthening, aerobic conditioning, and, if possible, protected ambulation. Programs to assure muscle strength to lift a limb against gravity should continue between orthopedic interventions using isotonic and aerobic conditioning. Swimming should be encouraged.

The goals of orthopedic surgery are to correct deformities for ambulation and to interrupt a cycle of fracture and re­fracture. The classic osteotomy procedere requires fixation with an intramedullary rod. The hardware currently in use includes telescoping rods (Bailey­Dubow [51] or Fassier­Duval rods) and nonelongating rods (Rush rods). Important considerations include selection of a rod with the smallest diameter suited to the situation to avoid cortical atrophy.

The secondary features of OI, including abnormal pulmonary function, hearing loss and basilar invagination are best managed in a specialized coordina­ted care program.

Conclusion

Osteogenesis imperfecta is a genetic disorder of connective tissue characterized by fragile bones and a susceptibility to fracture from mild trauma. The defects are either caused by primary defects in collagen type 1 or by proteins interacting with collagen type 1. The clinical range of this condition is extremely broad, ranging from lethal cases in the perinatal period to mild cases with only few fractures. Currently over 1500 dominant mutations in COL1A1 and ­COL1A2 have been identified [4]. Eleven types of OI have been classified, so far.

OI is diagnosed by radiologic and clinical findings or by collagen biochemical testing and DNA sequencing. The follow up of disease is made by radiological techniques, since bone turnover markers are not feasible for OI. DXA measurement might be a helpful tool for therapy monitoring. However, the DXA method has its limitations, especially in patients with OI. The HR­pQCT could be a helpful means in the evaluation of bone structure and quality.

Cyclic bisphosphonate therapy in OI treatment was firstly introduced in the late 1980’s. Numerous studies showed the positive effects of intravenous bisphosphonates on BMD, especially pamidronate and zoledronate acid. However, data on fracture risk and pain reduction are inconsistent. Oral bisphosphonate do not have significance in OI treatment. In future, osteoanabolic drugs, such as TPTD, sclerostin antibody and perhaps MSC could play a mayor role in the treatment of OI.