Radial Head Dislocation and Subluxation in Osteogenesis Imperfecta. Investigation performed at Shriners Hospital, Montreal, Quebec, Canada

Transcription

1 2694 COPYRIGHT Ó 2007 BY THE JOURNAL OF BONE AND JOINT SURGERY, INCORPORATED Raial Hea Dislocation an Subluxation in Osteogenesis Imperfecta By Alice Marcargent Fassier, MD, Frank Rauch, MD, Mehi Aarabi, MD, Chantal Janelle, MD, FRCS, an Francxois Fassier, MD, FRCS Investigation performe at Shriners Hospital, Montreal, Quebec, Canaa Backgroun: Upper limb eformity in chilren with osteogenesis imperfecta may substantially impair function. The aims of this retrospective work were to stuy the prevalence of raial hea malalignment (islocation or subluxation) in ifferent types of osteogenesis imperfecta an to ientify factors linke to it. Methos: We assesse 489 upper limbs from 254 patients (with a mean age of 9.6 years an incluing 130 female patients) who ha various types of osteogenesis imperfecta. Raiographs representing a single time-point for each patient were assesse for the presence an irection of raial hea malalignment an associate abnormalities (ysplasia of the capitellum or of the raial hea or neck, calcification of the interosseous membrane, or raioulnar synostosis). Deformations of the humerus, raius, an ulna were assesse with regar to location, irection, an magnitue. The forearm range of motion in pronation an supination an the han grip force were measure in a subset of patients. Results: We observe raial hea islocation or subluxation in forty-four an thirty-nine upper extremities, respectively. The frequency of raial hea malalignment was significantly higher in type-v osteogenesis imperfecta (86%) than in the other types (0% to 29%) (p < 0.001). Dysplasia of the humeral capitellum, raial hea, or raial neck was associate with malalignment in all types of osteogenesis imperfecta, with the exception of capitellum ysplasia in type V. Malalignment in type V was associate with calcification of the interosseous membrane, an abnormality that was specific for type V. In the other osteogenesis imperfecta types, malalignment was commonly linke with raial an ulnar eformation an was associate with ecrease forearm range of motion in supination an pronation an a lower grip force. Conclusions: Raial hea malalignment is common in osteogenesis imperfecta, especially in type V. Malalignment is associate with bowing characteristics an impaire function of the upper limb. These finings may provie support for surgical correction of raial an ulnar bowing in selecte patients with osteogenesis imperfecta. Level of Evience: Prognostic Level II. See Instructions to Authors for a complete escription of levels of evience. Osteogenesis imperfecta is a heritable isorer characterize by bone fragility, short stature, an bone eformities associate with typical extraskeletal manifestations such as entinogenesis imperfecta. It is most often cause by mutations affecting collagen type I 1. Disease severity ranges from intrauterine fractures an perinatal lethality to very mil forms with rare fractures. Sillence et al. separate osteogenesis imperfecta into four types 2. Type I inclues patients with a mil noneforming form an a normal or milly short stature. Type II is a perinatally lethal form relate to multiple rib an long bone fractures at birth. Type III is the most severe form for surviving chilren an is associate with progressive bone eformities an scoliosis. Type IV is a moerate form an inclues patients who o not conform to the previous categories. From this last an rather heterogeneous group of patients, three aitional types of osteogenesis imperfecta (types V, VI, an VII) have recently been elineate, on the basis of the absence of collagen type-i mutations an the specific clinical an histologic features of the bones. Type-V osteogenesis imperfecta is characterize by hyperplastic callus Disclosure: The authors i not receive any outsie funing or grants in support of their research for or preparation of this work. Neither they nor a member of their immeiate families receive payments or other benefits or a commitment or agreement to provie such benefits from a commercial entity. No commercial entity pai or irecte, or agree to pay or irect, any benefits to any research fun, founation, ivision, center, clinical practice, or other charitable or nonprofit organization with which the authors, or a member of their immeiate families, are affiliate or associate. J Bone Joint Surg Am. 2007;89: oi: /jbjs.f.01287

2 N D 2695 T HE J OURNAL OF B ONE &JOINT S URGERY JBJS. ORG V OLUME 89-A UMBER 12 ECEMBER 2007 R ADIAL H EAD D ISLOCATION AND S UBLUXATION IN O STEOGENESIS I MPERFECTA TABLE I Classification of the Types of Osteogenesis Imperfecta Type Severity Characteristics I Mil Normal stature, few or no bone eformities, vertebral fractures, triangular face, blue sclera, an able to walk II Perinatally lethal Multiple fractures at birth III Severe Short stature, scoliosis, triangular face, grayish sclera, an may be able to walk with ai IV Moerate Short stature, scoliosis, white sclera, an able to walk but may require ai V Moerate Moerately short stature, hyperplastic callus, ossification of interosseous membrane of the forearm an leg leaing to limite mobility, an white sclera VI Moeratesevere Moerately short stature, frequent fractures, vertebral compression, an white sclera VII Moerate Milly short stature, short humeri an femora, coxa vara, an white sclera formation, calcification of the interosseous membrane of the forearm an/or the leg, an an irregular pattern of bone lamellation seen uner polarize light microscopy 3. The iagnosis of type-vi osteogenesis imperfecta is base on specific histologic finings in the bone, incluing a mineralization efect affecting bone but not cartilage tissue 4.TypeVIIisa moerately severe form of osteogenesis imperfecta with short humeri, short femora, an coxa vara 5. A recent stuy foun that type VII is cause by a mutation in the CRTAP gene 6. Table I summarizes the current classification of osteogenesis imperfecta. Deformities of the upper extremity in osteogenesis imperfecta have been stuie far less frequently than those of the lower limbs, probably because they are less frequent an have less obvious functional implications. However, Amako et al. showe that upper limb eformities in chilren with osteogenesis imperfecta are not only a cosmetic problem but may substantially impair functional activities of aily living 7. The aims of the present retrospective stuy were to investigate the prevalence of raial hea malalignment in the ifferent types of osteogenesis imperfecta an to ientify factors linke to such malalignment. Materials an Methos Between January 1992 an June 2004, 376 patients with osteogenesis imperfecta were seen at least once in our institution. Patients were inclue in the present stuy if they coul be iagnose as having a specific type of osteogenesis imperfecta, i not have a recent or unheale upper limb fracture, an ha ha at least one raiographic assessment of the upper limb at our clinic. Patients in whom raial hea alignment was impossible to efine on the raiographs were exclue. Two hunre an fifty-four patients (130 female an 124 male patients) fulfille these inclusion criteria. The mean age was 9.5 years (range, 1.5 months to 23.5 years). Seventy-six patients (30%) ha type-i; sixty (23.5%), type-iii; eighty-six TABLE II Distribution of Upper Limbs in Stuy Patients Accoring to Age, Sex, Sie, an Type of Osteogenesis Imperfecta Type of Osteogenesis Imperfecta* I III IV V VI VII All Mean age (yr) Sex Sie Male Female Right Left Total Percentage of all types *A significant ifference was foun among the types of osteogenesis imperfecta with respect to age (p = 0.046) an sex istribution (p < 0.001).

3 2696 THE JOURNAL OF BONE & JOINT SURGERY JBJS.ORG V O L U M E 89-A N U M B E R 12 D E C E M B E R RADIAL HEAD DISLOCATION OSTEOGENESIS IMPERFECTA Fig. 1-A AND SUBLUXATION IN Fig. 1-B Anteroposterior (Fig. 1-A) an lateral (Fig. 1-B) raiographs of the left elbow of a patient with type-iv osteogenesis imperfecta, showing lateral subluxation of the raial hea. The straight line represents the longituinal axis of the proximal part of the raius. The otte line represents the contours of the normal capitellum. The cross represents the center of the capitellum. On the anteroposterior raiograph, the longituinal axis of the proximal part of the raius passes through the capitellum but not in the center, efining raial hea subluxation. (34%), type-iv; eighteen (7%), type-v; ten (4%), type-vi; an four (1.5%), type-vii osteogenesis imperfecta. The classification of patients accoring to the type of osteogenesis imperfecta was primarily base on clinical characteristics. However, in all patients with types V, VI, an VII, the iagnosis ha been confirme by the absence of collagen type-i mutations an specific finings on histologic analysis of the iliac bone. For each patient, we selecte raiographs from a single time-point, when ocumentation of the relevant skeletal areas was best. Anteroposterior an lateral raiographs of 489 upper extremities were available. Most of these raiographs were mae as part of routine skeletal surveys. As this is a hospitalbase stuy, the stuy group probably oes not provie a representative cross-section of the entire population of patients with osteogenesis imperfecta but most likely is skewe towar those who are more severely affecte. In some patients, raiographs from only one upper limb were available. For this reason, results are presente relative to the number of upper limbs rather than relative to the number of patients. The istribution of examine upper limbs accoring to age, sex, sie, an type of osteogenesis imperfecta is shown in Table II. Raiographic Assessment There are no establishe raiographic criteria to assess raial hea malalignment an associate abnormalities. Nevertheless, islocation is commonly efine as the complete isruption of a joint, whereas subluxation is a minor isruption of a joint in which some articular contact remains8. Accoring to the criteria efine by Sachar an Mih, we therefore istinguishe between raial hea islocation an raial hea subluxation9. The other raiographic criteria use in this stuy are relate to normal anatomical escriptions. Using anteroposterior an lateral raiographs of the humerus an forearm, we recore the presence an irection of raial hea islocation or raial hea subluxation, the

4 2697 THE JOURNAL OF BONE & JOINT SURGERY JBJS.ORG V O L U M E 89-A N U M B E R 12 D E C E M B E R RADIAL HEAD DISLOCATION OSTEOGENESIS IMPERFECTA Fig. 2-A AND SUBLUXATION IN Fig. 2-B Lateral (Fig. 2-A) an anteroposterior (Fig. 2-B) plain raiographs of the right elbow of a patient with type-v osteogenesis imperfecta, showing anterolateral islocation of the raial hea. The straight line represents the longituinal axis of the proximal part of the raius. The otte line represents the contours of the normal capitellum. The cross represents the center of the capitellum. On both anteroposterior an lateral raiographs, the longituinal axis of the proximal part of the raius oes not pass through the capitellum, efining raial hea islocation. presence of a ysplastic capitellum, the presence of a ysplastic raial hea or raial neck, an the presence of interosseous membrane ossification an/or raioulnar synostosis. A normally aligne raial hea was sai to be present when a line coincient with the longituinal axis of the proximal aspect of the raius passe through the center of the capitellum on both anteroposterior an lateral raiographs. If this line passe through the capitellum but not through its center, we classifie the fining as raial hea subluxation (Figs. 1-A an 1-B). If this line i not pass through the capitellum, we efine it as a raial hea islocation (Figs. 2-A an 2-B). Raial hea islocation an raial hea subluxation are groupe uner the term malaligne raial hea. The normal capitellum is a roune, convex projection, less than half a sphere, an forme of the anterior an inferior surfaces of the lateral part of the conyle of the humerus10 (Fig. 3-A). If the capitellum i not correspon to this escription, we efine it as ysplastic (Fig. 3-B). The normal raial hea is isc-shape, an its articular surface is a shallow cup for articulation with the capitellum of the humerus. The articular circumference of the hea is smooth an is wiest meially, where it articulates with the raial notch of the ulna. The neck of the raius is the constricte part istal to the raial hea an is overhung by it, especially on the lateral sie10 (Fig. 3-A). If the raial hea an/or the raial neck i not correspon to this escription, we efine it as a ysplastic raial hea or raial neck (Fig. 3-B). In young patients in whom the capitellum or the raial hea ossification centers ha not yet appeare, only the clearly ientifiable abnormalities were note. The normal interosseous membrane originates approximately 3 cm istal to the raial tuberosity an extens to the wrist11. If the interosseous membrane was not raiolucent, we efine it as calcifie (Fig. 4). A raioulnar synostosis is a fusion between those two bones11 (Fig. 5).

5 2698 THE JOURNAL OF BONE & JOINT SURGERY JBJS.ORG V O L U M E 89-A N U M B E R 12 D E C E M B E R RADIAL HEAD DISLOCATION OSTEOGENESIS IMPERFECTA AND SUBLUXATION IN TABLE III Frequency of Raial Hea Dislocation an Subluxation by Type of Osteogenesis Imperfecta Type of Osteogenesis Imperfecta* I (N = 142) III (N = 116) IV (N = 166) V (N = 36) VI (N = 22) VII (N = 7) All (N = 489) No. (%) of limbs with raial hea islocation 2 (1) 10 (9) 11 (7) 21 (58) 0 (0) 0 (0) 44 (9) No. (%) of limbs with raial hea subluxation 1 (<1) 15 (13) 11 (7) 10 (28) 0 (0) 2 (29) 39 (8) All 3 (2) 25 (22) 22 (13) 31 (86) 0 (0) 2 (29) 83 (17) *The frequency of raial hea islocation an raial hea subluxation iffere significantly among the types of osteogenesis imperfecta (p < 0.001). The location (proximal, mile, an istal thir), irection, an magnitue of humeral, raial, an ulnar bowing eformities were recore for each type of osteogenesis imperfecta. If bowing occurre in several locations or ha several irections, we classifie the fining as multiple bowing. We Fig. 3-A measure the angle between the long axis of the proximal an istal parts of the bone for both the anteroposterior an the lateral raiograph (Figs. 6-A an 6-B). The magnitue of the eformity was efine by the largest angle measure in either raiograph7. Fig. 3-B Anteroposterior raiographs of the right an left elbows of a patient with type-v osteogenesis imperfecta, showing a normal (Fig. 3-A) an a ysplastic (Fig. 3-B) capitellum, raial hea, an raial neck. The otte line on the lateral sie of the istal part of the humerus represents the contours of the capitellum. The otte line on the proximal part of the raius represents the contours of the raial hea an raial neck.

6 N D 2699 T HE J OURNAL OF B ONE &JOINT S URGERY JBJS. ORG V OLUME 89-A UMBER 12 ECEMBER 2007 R ADIAL H EAD D ISLOCATION AND S UBLUXATION IN O STEOGENESIS I MPERFECTA Fig. 4 Raiograph of the left forearm of a patient with type-v osteogenesis imperfecta showing a calcifie interosseous membrane. Clinical Evaluation The results of clinical examinations were inclue in this stuy if they ha been performe within six months of the raiograph. Measurements of the range of motion in pronation an supination as well as han grip force were obtaine by experience occupational therapists. Three measurements of grip force were performe in the same session, an the mean of the three was note. Grip force results were converte into agespecific z-scores with use of publishe reference ata 12. Statistical Analysis Results of iscrete measures were isplaye as frequencies, an the significance between the groups was assesse by the chisquare test. Continuous variables were assesse as means an stanar eviations, an comparisons between the means in two groups were performe with a Stuent t test or a Mann- Whitney U test, as appropriate. Comparison between means in more than two groups was performe by a one-way analysis of variance. Os ratios erive from logistic regression analysis were use to evaluate factors that were inepenently associate with raial hea malalignment. A p value of <0.05 was consiere significant. Results In this series of 489 upper limbs, we observe eighty-three limbs (17%) with raiocapitellar joint malalignment (Table III). Raial hea islocation was present in 9% an raial hea subluxation was present in 8% of the limbs. The frequency of raial hea malalignment iffere among osteogenesis imperfecta types, ranging from 0% in type VI to 86% in type V. The frequency of raial hea islocation was similar to that of raial hea subluxation for all types except for type V an type VII. In type V, raial hea islocation was more prevalent (58% of the upper limbs in the group) than raial hea subluxation (28%). This was opposite to the finings in the seven limbs in the type-vii group, in which the raial hea was more often subluxate (two upper limbs) than islocate (no upper limb). Logistic regression analyses reveale that the prevalence of raial hea islocation increase with age, regarless of the type of osteogenesis imperfecta an the sex of the patient (Table IV). Raial hea islocation was also more prevalent in types III, IV, an V than in type I, which was selecte as the reference type for the purpose of the logistic regression analysis. The type of osteogenesis imperfecta, but not age, was also inepenently associate with raial hea subluxation (Table IV). As raial hea malalignment seeme to be more frequent in type-v osteogenesis imperfecta than in the other types, we attempte to characterize the ifferences among the types more closely. The irection of islocation was significantly ifferent between type V an the other types (p < 0.001). We foun that in type-v osteogenesis imperfecta, nineteen (90%) of twentyone limbs with raial hea islocation ha islocate in an anterior, lateral, or anterolateral irection (Figs. 2-A an 2-B), whereas this ha occurre in only two (9%) of the twenty-three limbs with raial hea islocation in the other types of osteogenesis imperfecta. For types other than type V, posterior an posterolateral islocations were most often seen, characterizing sixteen (70%) of twenty-three such limbs, whereas such islocations occurre in only two (10%) of twenty-one limbs with raial hea islocation in type-v osteogenesis imperfecta. No pure meial but five anteromeial or posteromeial raial hea islocations (Figs. 7-A an 7-B) were observe, an all of them were in limbs with osteogenesis imperfecta of a type other than type V. In type-v osteogenesis imperfecta, all raial hea subluxations were anterior, lateral, or anterolateral, following the same irections most frequently seen in islocations. For other types of osteogenesis imperfecta, twenty-four (83%) of

7 N D 2700 T HE J OURNAL OF B ONE &JOINT S URGERY JBJS. ORG V OLUME 89-A UMBER 12 ECEMBER 2007 R ADIAL H EAD D ISLOCATION AND S UBLUXATION IN O STEOGENESIS I MPERFECTA TABLE IV Effect of Age, Sex, an Type of Osteogenesis Imperfecta on the Frequency of Raial Hea Dislocation an Subluxation Accoring to Logistic Regression Analysis Raial Hea Dislocation* Raial Hea Subluxation Yes No P Value Os Ratio Yes No P Value Os Ratio Mean age (yr) < Sex (F/M) 25/19 228/ /18 232/ Type of osteogenesis imperfecta (no. of limbs) All types < III IV V < < VI VII *A significant association was foun between raial hea islocation an age an type of osteogenesis imperfecta; no significant association was foun between raial hea islocation an sex. A significant association was foun between raial hea subluxation an type of osteogenesis imperfecta; no significant association was foun between raial hea subluxation an age an sex. P values inicate the significance of the association after ajustment for the other two explanatory variables. twenty-nine limbs with raial hea subluxation were lateral, posterior, or posterolateral. However, with the numbers stuie, this ifference in the irection of subluxation was not foun to be significant (p = 0.053). No meial or anteromeial subluxations an only one posteromeial subluxation were observe in osteogenesis imperfecta types other than type V. More than 95% of the limbs coul be assesse for an abnormal calcification of the interosseous membrane or a ysplastic appearance of the capitellum, raial hea, or raial neck. A ysplastic capitellum was significantly associate with malalignment in all types of osteogenesis imperfecta (p = for raial hea islocation in type V, an p < for raial Fig. 5 Lateral raiograph of the right elbow of a patient with type-i osteogenesis imperfecta showing a raioulnar synostosis.

8 2701 THE JOURNAL OF BONE & JOINT SURGERY JBJS.ORG V O L U M E 89-A N U M B E R 12 D E C E M B E R RADIAL HEAD DISLOCATION OSTEOGENESIS IMPERFECTA Fig. 6-A AND SUBLUXATION IN Fig. 6-B Anteroposterior (Fig. 6-A) an lateral (Fig. 6-B) raiographs of the right forearm of a patient with type-v osteogenesis imperfecta, emonstrating the metho use to measure the magnitue of the ulnar eformity. The lines represent the axes of the proximal an istal parts of the bone for both views. a is the ulnar eformity angle on the anteroposterior raiograph. b is the ulnar eformity angle on the lateral raiograph. The magnitue of ulnar eformity in this case is b, because it is the larger angle measure. hea islocation an raial hea subluxation in the other types of osteogenesis imperfecta), except for raial hea subluxation in type V, which may be attribute to the small number of limbs with raial hea subluxation in patients with type-v isease. A ysplastic raial hea or raial neck was associate with malalignment in all types of osteogenesis imperfecta (p < 0.02 for both type V an the other types). Calcification of the interosseous membrane was foun only in type V an was associate with both raial hea islocation (p = 0.01) an raial hea subluxation (p = 0.03) in this group. Raioulnar synostosis was not associate with malalignment in type V, an was linke only to raial hea islocation in the other types of osteogenesis imperfecta (p < 0.001). The irection of bowing eformities in the humerus an forearm bones coul be evaluate in 87% of the limbs. In typev osteogenesis imperfecta, malalignment of the raial hea was associate only with ulnar bowing (p = 0.02 for raial hea islocation an p = 0.03 for raial hea subluxation) but not with raial bowing. In this type, eight of nineteen limbs with raial hea islocation ha ulnar bowing in multiple irec-

9 2702 THE JOURNAL OF BONE & JOINT SURGERY JBJS.ORG V O L U M E 89-A N U M B E R 12 D E C E M B E R RADIAL HEAD DISLOCATION OSTEOGENESIS IMPERFECTA Fig. 7-A AND SUBLUXATION IN Fig. 7-B Lateral (Fig. 7-A) an anteroposterior (Fig. 7-B) raiographs of the right elbow of a patient with type-iv osteogenesis imperfecta, showing a posteromeial islocation of the raial hea. tions, an seven of nine limbs with raial hea subluxation ha ulnar bowing in a lateral or anterior irection. The magnitue of the ulnar bowing in the limbs with raial hea islocation (mean, 23 ) was significantly higher (p = 0.02) than that in the upper limbs without raial hea islocation (mean, 12 ). In osteogenesis imperfecta other than type V, the irection an location of raial an ulnar bowing but not of humeral bowing was relate to malalignment. However, the magnitue of the raial bowing was not linke with malalignment. The magnitue of the ulnar bowing was significantly higher in the presence of raial hea islocation (26 ) an raial hea subluxation (19 ) than in the absence of raial hea malalignment (11 ) (p < 0.01). Aitionally, we assesse the relationship between raial hea malalignment an functional stuies, namely, han grip force an forearm motion in pronation an supination. Sufficient numbers of stuies were available only for patients with type-iii, IV, an V osteogenesis imperfecta. Information was available for 25% to 47% of the eighty-three limbs affecte by raial hea malalignment, epening on the criterion use. In osteogenesis imperfecta other than type V, the range of motion in pronation an supination was significantly lower when raial hea islocation was present (p < 0.001). This range of motion was not linke to raial hea islocation in type-v osteogenesis imperfecta an was not linke to raial hea subluxation regarless of type. The presence of raial hea islocation was associate with lower grip force, both in type V (an average grip force z-score of 3.8 with raial hea islocation compare with 2.2 without islocation; p = 0.008) an in type-iii or IV osteogenesis imperfecta (an average grip force z-score of 4.0 with raial hea islocation compare with 2.6 without islocation; p = 0.05). Discussion n this series of patients with osteogenesis imperfecta, malalignment was present in a substantial number of patients an the prevalence of malalignment epene on the type of osteogenesis imperfecta. Raial hea malalignment was particularly frequent in type V. In the so-calle classic types I through IV, the frequency of raial hea islocation seeme to mirror the severity of the isease (type III > type IV > type I). Previous work from our institution ha expane the Sillence classification of osteogenesis imperfecta by elineating three aitional types of the isorer (types V, VI, an VII)1. The present stuy highlights the utility of the expane classification, as it shows that type V is also a istinct entity I

10 N D 2703 T HE J OURNAL OF B ONE &JOINT S URGERY JBJS. ORG V OLUME 89-A UMBER 12 ECEMBER 2007 R ADIAL H EAD D ISLOCATION AND S UBLUXATION IN O STEOGENESIS I MPERFECTA with regar to raial hea malalignment. The most obvious specific characteristic of raial hea malalignment in type V is its association with calcification of the interosseous membrane. This calcification was note in the first escription of type-v osteogenesis imperfecta by Glorieux et al., who also note that the frequency of raial hea islocation seeme unusually high in this isorer 3, an impression confirme by the present stuy. In aition, we foun that an anterior an/or lateral irection of malalignment was typical of type V. Dysplastic eformation of the capitellum an raial hea or neck was also frequently encountere in type V but was not specific for this type. Such ysplastic eformations are also well known from the literature on congenital raial hea islocation 9, Raial hea malalignment in osteogenesis imperfecta types other than type V preominantly occurre in a posterior an posterolateral irection, also the usual irection of islocation in congenital raial hea islocation 16. The association between raioulnar synostosis an raial hea islocation is another similarity to congenital raial hea islocation 13.Bowing of both forearm bones was frequently encountere in patients with raial hea malalignment, which is not surprising as fractures are more frequent in the more severe types of osteogenesis imperfecta, in turn leaing to a higher prevalence of malalignment. In the presence of malalignment, ulnar bowing was note preominantly in a posterior or posterolateral irection. Thus, raial hea malalignment an ulnar bowing commonly ha the same irection. It is still unclear whether this association reflects a causal relationship or is a mere coincience. The present ata inicate that patients with raial hea islocation ha functional impairment with regar to grip force an the forearm range of motion in pronation an supination. Similar finings were reporte in a small group of patients with congenital raial hea islocation 15. The factors causing raial hea malalignment in osteogenesis imperfecta were not irectly aresse in this stuy, but our observations may nevertheless give rise to hypotheses about the etiology of this problem. Schematically, raial hea malalignment might be ue to congenital, evelopmental, or traumatic factors. A traumatic etiology is thought to be rare in general, but it might be more frequent in patients with osteogenesis imperfecta, who can have high fracture rates 17,18. However, this possibility coul not be investigate on the basis of the present stuy material. In our series, we i not fin the so-calle ulnar bow sign (anterior bow of the ulna) escribe by Lincoln an Mubarak, which can help to ientify a Monteggia-like lesion or islocation 18. Posterior or posterolateral ulnar bowing was the only eformity significantly linke to raial hea islocation in osteogenesis imperfecta other than type V (p < 0.01). Progressive long-bone bowing is probably a major factor leaing to subluxation or islocation in such patients, whereas calcification of the interosseous membrane likely is the most critical factor in type V. Prospective stuies are neee to she more light on these etiological aspects. Although this is by far the largest stuy on raial hea malalignment in osteogenesis imperfecta, the present analysis has a number of limitations. Most importantly, this is a retrospective review of raiographs an patient charts. This may have introuce some selection bias, as raiographic ocumentation of the upper limb is probably better in patients who presente with clinical problems in that anatomic area. Nevertheless, most of the patients more severely affecte with osteogenesis imperfecta ha regular skeletal surveys of all extremities, regarless of clinical manifestations. A retrospective an cross-sectional stuy such as the present one is not a suitable basis for specific treatment recommenations. Nevertheless, we note that multiple locations an irections of forearm eformities an an ulnar bowing magnitue of >20 were frequently associate with raial hea islocation in patients with osteogenesis imperfecta other than those with type V. It therefore appears reasonable to hypothesize that the appearance of raial hea islocation coul be prevente by surgical correction of such eformities. In type-v osteogenesis imperfecta, such an approach appears less promising, as the calcification of the interosseous membrane seems to be an important etiological factor. Further stuies are necessary to test whether straightening of the forearm bones inee prevents raial hea islocation an improves upper extremity function. n NOTE: The authors thank Guylaine Bear for illustrations appearing on the figures an Kathleen Montpetit, occupational therapist, for sharing her experience with patients with osteogenesis imperfecta. Alice Marcargent Fassier, MD Peiatric Orthopaeics Clinical Fell, McGill University, 1529 Cear Avenue, Montreal, QC H3G 1AG, Canaa. aress: amarcargent@shriners.mcgill.ca Frank Rauch, MD Chantal Janelle, MD, FRCS Francxois Fassier, MD, FRCS Genetics Unit (F.R.), Peiatric Orthopaeics (C.J. an F.F.), an Han an Upper Limb Surgery (C.J.), Shriners Hospital for Chilren, 1529 Cear Avenue, Montreal, QC H3G 1A6, Canaa. aress for F. Rauch: frauch@shriners.mcgill.ca. aress for C. Janelle: cjanelle@shriners.mcgill.ca. aress for F. Fassier: ffassier@shrinenet.org Mehi Aarabi, MD Department of Orthopaeic Surgery, Toronto East General Hospital, 825 Coxwell Avenue, Toronto, ON M4C 3E7, Canaa. aress: rmaarabi@yahoo.ca References 1. Rauch F, Glorieux FH. Osteogenesis imperfecta. Lancet. 2004;363: Sillence DO, Senn A, Danks DM. Genetic heterogeneity in osteogenesis imperfecta. J Me Genet. 1979;16: Glorieux FH, Rauch F, Plotkin H, War L, Travers R, Roughley P, Lalic L, Glorieux DF, Fassier F, Bishop NJ. Type V osteogenesis imperfecta: a new form of brittle bone isease. J Bone Miner Res. 2000;15:

11 N D 2704 T HE J OURNAL OF B ONE &JOINT S URGERY JBJS. ORG V OLUME 89-A UMBER 12 ECEMBER 2007 R ADIAL H EAD D ISLOCATION AND S UBLUXATION IN O STEOGENESIS I MPERFECTA 4. Glorieux FH, War LM, Rauch F, Lalic L, Roughley PJ, Travers R. Osteogenesis imperfecta type VI: a form of brittle bone isease with a mineralization efect. J Bone Miner Res. 2002;17: War LM, Rauch F, Travers R, Chabot G, Azouz EM, Lalic L, Roughley PJ, Glorieux FH. Osteogenesis imperfecta type VII: an autosomal recessive form of brittle bone isease. Bone. 2002;31: Morello R, Bertin TK, Chen Y, Hicks J, Tonachini L, Monticone M, Castagnola P, Rauch F, Glorieux FH, Vranka J, Bachinger HP, Pace JM, Schwarze U, Byers PH, Weis M, Fernanes RJ, Eyre DR, Yao Z, Boyce BF, Lee B. CRTAP is require for prolyl 3-hyroxylation an mutations cause recessive osteogenesis imperfecta. Cell. 2006;127: Amako M, Fassier F, Hamy RC, Aarabi M, Montpetit K, Glorieux FH. Functional analysis of upper limb eformities in osteogenesis imperfecta. J Peiatr Orthop. 2004;24: Greenspan A. Orthopeic imaging: a practical approach. 4th e. Philaelphia: Lippincott, Williams an Wilkins; Raiologic evaluation of trauma; p Sachar K, Mih AD. Congenital raial hea islocations. Han Clin. 1998;14: Osteology. In: Williams PL, Warwick R. Gray s anatomy. 36th e. Philaelphia: WB Sauners; p Resnick D, Kransorf M. Bone an joint imaging. 3r e. Philaelphia: Elsevier Sauners; Articular anatomy an histology; p Rauch F, Neu CM, Wassmer G, Beck B, Rieger-Wettengl G, Rietschel E, Manz F, Schoenau E. Muscle analysis by measurement of maximal isometric grip force: new reference ata an clinical applications in peiatrics. Peiatr Res. 2002;51: Almquist EE, Goron LH, Blue AI. Congenital islocation of hea of raius. J Bone Joint Surg Am. 1969;51: Echtler B, Burckhart A. Isolate congenital islocation of the raial hea. Goo function in 4 untreate patients after years. Acta Orthop Scan. 1997;68: Reichenbach H, Hörmann D, Theile H. Hereitary congenital posterior islocation of raial heas. Am J Me Genet. 1995;55: Maram-Bey T, Ger E. Congenital raial hea islocation. J Han Surg Am. 1979;4: Kemnitz S, De Schrijver F, De Smet L. Raial hea islocation with plastic eformation of the ulna in chilren. A rare an frequently misse conition. Acta Orthop Belg. 2000;66: Lincoln TL, Mubarak SJ. Isolate traumatic raial-hea islocation. J Peiatr Orthop. 1994;14:454-7.