Note: This copy is for your personal, non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, cont

Transcription

1 Note: This copy is for your personal, non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at ORIGINAL RESEARCH n PEDIATRIC IMAGING Sarah D. Bixby, MD Delma Y. Jarrett, MD Travis Matheney, MD Patrick Johnston, MMath James Kasser, MD Paul K. Kleinman, MD Unilateral Subtalar Coalition: Contralateral Sustentaculum Tali Morphology1 Purpose: Materials and Methods: To measure and compare the dimensions of the sustentaculum tali (ST ) in pediatric patients with unilateral subtalar coalition to determine if the contralateral side has altered morphology. The institutional review board approved this study, which was HIPAA compliant. Informed consent was waived. Radiology records were reviewed to identify 35 patients with unilateral subtalar coalition undergoing computed tomography (CT) (21 male, 14 female; mean age, years) and 33 control patients with triplane fracture (21 male, 12 female; mean age, years). CT images were reviewed, and reformatted images through the subtalar joint (on the side opposite the coalition) were created. Anteroposterior measurements of the middle facet (MF) and the ST were recorded by two observers with electronic calipers. The MF/ST ratio and the ST length extending posterior to the MF were calculated. Measurements between groups were compared by using Wald tests based on linear regressions. Inter- and intraobserver variabilities were determined by means of a component of variance model. Results: Conclusion: The lengths of the MF and ST for the study and control groups were and mm ( P,.001) and and mm ( P =.053), respectively. Intra- and interobserver correlations for both measurements were 0.94 and 0.92 and 0.86 and 0.77, respectively. MF/ST ratio was 0.54 versus 0.76 ( P,.001), and ST length extending posterior to the MF was versus 5.24 mm ( P,.001). The MF is smaller and the ST extends further beyond the MF in patients with a contralateral subtalar coalition than in control patients. The morphology of the ST may provide insight into the origins and development of coalitions. q RSNA, From the Department of Radiology (S.D.B., D.Y.J., P.K.K.), Department of Orthopedic Surgery (T.M., J.K.), and Clinical Research Program (P.J.), Children s Hospital Boston, 300 Longwood Ave, Boston, MA Received November 24, 2009; revision requested January 8, 2010; fi nal revision received May 17; accepted June 2; fi nal version accepted July 10. Address correspondence to S.D.B. ( sarah.bixby@childrens.harvard.edu ). q RSNA, radiology.rsna.org n Radiology: Volume 257: Number 3 December 2010

2 Subtalar coalition is a common cause of repeated ankle sprains, peroneal muscle spasm, anterior ankle and hindfoot pain, and restricted subtalar motion in children and adolescents ( 1 3 ). Previously thought to affect less than 1% of the population ( 4 ), the prevalence may be as high as 13% if asymptomatic coalitions also are considered ( 5,6 ). Although the abnormality may be suspected clinically ( 7 ), imaging studies can be used to confirm the diagnosis ( 8 12 ) and help to define the type (osseous vs nonosseous) ( 11 ) and extent of coalition for preoperative planning ( 13 ). Talocalcaneal (subtalar) coalition is the second most common type of tarsal coalition after calcaneonavicular coalition ( 6,11 ). The middle facet (MF) of the subtalar joint is considered the most common location for a subtalar coalition to occur ( 14,15 ). We have noted that some patients with a unilateral subtalar coalition appear to have a contralateral sustentaculum tali (ST) that is larger and extends more posteriorly than expected. Elucidation of a primary morphologic abnormality of the ST associated with subtalar coalition may provide a clue to the etiology of this important condition. Thus, the purpose of this study was to measure and compare the dimensions of the ST in pediatric patients with unilateral subtalar coalition to determine if the contralateral side has altered morphology. Advances in Knowledge n The anteroposterior dimension of the middle facet (MF) is smaller on the apparently normal ankle opposite a symptomatic subtalar coalition than in a control population. n The dimension of the unaffected sustentaculum tali (ST) is larger in patients with coalition than in control subjects, and this difference is largely accounted for by abnormal posterior extension of the ST beyond a small MF. Materials and Methods Patient Population The study was approved by the institutional review board and was compliant with the Health Insurance Portability and Accountability Act. Because this was a retrospective cross-sectional review, informed consent was waived. We reviewed computed tomographic (CT) reports from January 1998 to June 2008 to identify pediatric patients aged 0 to 18 years who underwent CT of the lower extremity with notation of coalition (keyword coalition ) as the indication for the examination. This review generated a list of 155 patients. Of these patients, 91 had talocalcaneal (subtalar) coalition, 64 had calcaneonavicular coalition, one had talonavicular coalition, and one had cuboid-cuneiform coalition. CT images for the patients with talocalcaneal coalition were reviewed by one radiologist (S.D.B., 3 years of experience) to confirm the presence of subtalar coalition and to identify patients with unilateral coalition. Of the patients with talocalcaneal coalition, 46 had bilateral and 45 had unilateral coalition. Of the 45 patients with unilateral coalition, 35 had digitally acquired images available for review on a picture archiving and communication system (PACS) workstation (21 male, 14 female; mean age, years [standard deviation]). The electronic medical records were reviewed to confirm that the laterality of the patients symptoms corresponded to the site of the coalition at CT on the basis of the orthopedic clinic note before CT. In a separate review, we searched CT records from January 1998 to June 2008 for the words triplane fracture as the indication for the examination. Images from 114 CT scans were reviewed by one radiologist (S.D.B.) Implication for Patient Care n Pediatric patients with a unilat- eral subtalar coalition demonstrate morphologic abnormalities on the normal side that may indicate a potential for future symptoms in that ankle. to identify studies in which the CT of the ankle extended through the level of the entire talocalcaneal joint. This review generated a total of 33 patients (21 male, 12 female; mean age, years ). These patients comprised the control population ( Fig 1 ). Image Acquisition The CT techniques varied across the study period. Twenty-four studies were performed with a single detector row scanner (HiSpeed Advantage CT; GE Medical Systems, Milwaukee, Wis) with images acquired helically at 1-mm section thickness. Twenty-eight studies were performed with a 16 detector row scanner (LightSpeed Pro VCT; GE Medical Systems), and two studies were performed with a 32 detector row scanner (LightSpeed Pro VCT; GE Medical Systems). These images were acquired helically at mm section thickness and were reconstructed at 1.25-mm thickness by using a bone algorithm. Two studies were performed helically (Somatom Sensation 40 scanner; Siemens Medical Systems, Erlangen, Germany) with a section thickness of 0.75 mm reconstructed to 1.0 mm by using a bone algorithm, and 12 studies were performed helically (Somatom Sensation 64 scanner; Siemens Medical Systems) with a section thickness of 0.60 mm reconstructed to 1.0 mm by using a bone algorithm. For patients suspected of having tarsal coalition, both feet were placed in the gantry with the knees and hips bent and the Published online before print /radiol Radiology 2010; 257: Abbreviations: MF = middle facet PACS = picture archiving and communication system ST = sustentaculum tali Author contributions: Guarantor of integrity of entire study, S.D.B.; study concepts/ study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, S.D.B., T.M., P.K.K.; clinical studies, S.D.B., T.M., J.K., P.K.K.; statistical analysis, P.J.; and manuscript editing, S.D.B., P.J., J.K., P.K.K. Authors stated no fi nancial relationship to disclose. Radiology: Volume 257: Number 3 December 2010 n radiology.rsna.org 831

3 Figure 1 Figure 2 Figure 2: Three-dimensional reconstructed CT image obtained by using a bone algorithm in 13-year-old adolescent boy demonstrates a nonosseous subtalar coalition (arrow) posterior to the MF. The MF is otherwise normal (arrowhead). Figure 1: Flow diagram demonstrates the patient selection method. The radiology database was queried retrospectively for all CT scans of the lower extremities (CTLE) obtained between January 1998 and June 2008 with keywords coalition and triplane fracture as indication for examination. The study and control populations were then compiled as outlined. feet plantar flexed on the table surface. For patients with triplane fractures, the knees and hips were extended, and the ankles were maintained in neutral position. Images were acquired perpendicular to the long axis of the foot in those suspected of having coalition and parallel to the long axis of the foot for patients with triplane fracture. Image Interpretation Each set of CT images was reviewed by one radiologist (P.K.K., 30 years of experience) to confirm signs of talocalcaneal coalition on a PACS workstation. This reader did not participate in the subsequent quantitative assessment. CT criteria for coalition were based on findings previously established in the literature, including demonstration of an osseous bar between the talus and calcaneus ( 16,17 ), or the presence of articular narrowing with cystic joint irregularity or subchondral sclerosis ( Fig 2 ) ( 11,17 ). Multiplanar reformatting of each CT scan was performed by one radiologist (S.D.B.) using software (Voxar 3D; Voxar, Edinburgh, Scotland). Reformatted planes were aligned with respect to the axis of the ST ( Fig 3 ). These reformatted images were sent to a PACS workstation for review. The MF and the ST were measured anteroposteriorly on sagittal and transverse (footprint) images by the same radiologist using electronic calipers. This reader obtained two sets of measurements for assessment of intrareader variability, with each set of measurements for each patient obtained at separate visits separated in time by 4 6 weeks to avoid recall bias. A second reader (D.Y.J., 2 years of experience) obtained the same measurements in each patient, in random order, for assessment of interreader variability. Measurements in patients with unilateral coalition were obtained only on the side contralateral to the coalition. A ratio of the length of the MF over the length of the ST (MF/ST) was calculated and recorded for each measurement. A measurement of the portion of the ST extending posterior to the MF was determined by subtracting the MF measurement from the ST measurement ( Fig 4 ). Statistical Analysis With respect to covariates, the groups were compared by means of linear regression for age and CT section thickness and logistic regression for sex. Intraobserver reliability between two measurements from a single reader was estimated by means of a components of variance model. This method also was used to estimate interobserver variability between the two readers. Groups were compared with respect to outcome measurements by using Wald tests based on linear regressions adjusting for age, sex, and CT section thickness. Results Groups were not significantly different with respect to age, sex, or CT section thickness ( Table 1 ). The measurements for the MF, ST, MF/ST ratio, and ST length extending posterior to the MF for each group are demonstrated in Table 2. Intraobserver correlations 832 radiology.rsna.org n Radiology: Volume 257: Number 3 December 2010

4 Figure 3 Figure 4 between measurements by the same reader were 0.94 and 0.92 for the MF and ST, respectively. Given this high level of agreement, only the first measurement for each patient was used for statistical analysis. Interobserver correlations between measurements by two readers were 0.86 and 0.77 for the MF and ST, respectively. As outlined in Table 2, the MF was smaller, the ST was larger, and MF/ST ratio was smaller in the patients with a contralateral coalition than in patients with triplane fracture. The mean measurement of the ST extending posterior to the MF was 2.09 times greater in patients with unilateral coalition than in patients with Figure 3: Multiplanar reformatted CT images obtained by using a bone algorithm in 13-year-old adolescent girl include (a) transverse, (b) coronal, and (c) sagittal images aligned with reference to the ST (arrows). Anteroposterior measurements of the ST and MF were acquired from transverse and sagittal reformatted images. triplane fracture (10.97 vs 5.24 mm) ( Fig 5 ). Discussion Tarsal coalition generally is considered an uncommon condition ( 4 ), although cadaveric and more recent imaging studies indicate that its prevalence may be as high as 13% ( 5,6 ). Advanced imaging techniques and improved recognition of secondary signs of coalition also have led to the increase in diagnosis ( 9 ). Although indirect indicators of subtalar coalition, such as the C sign ( 18 ) and the talar beak ( 19 ), on conventional radiographs are well described in the literature, many subtalar coalitions are radiographically occult largely because of the complex orientation of the subtalar joint ( 16 ). CT is more sensitive for the detection of subtalar coalition, particularly with multiplanar reformations generated with isotropic helical acquisitions ( 12 ). Despite the superior capacity of CT to define bone anatomy, nonosseous Figure 4: Measurement method for determining MF and ST length. Sagittal reformatted CT image obtained by using a bone algorithm in 13-year-old adolescent girl with a contralateral subtalar coalition. The portion of the ST posterior to the MF is determined by subtracting the MF measurement from the ST measurement. subtalar coalitions still may remain undiagnosed in cases in which an osseous deformity is not present ( 11 ). In one study ( 20 ), fibrous subtalar coalitions occurred exclusively in the posterior part of the MF or posterior to the ST. These coalitions were not identified easily on preoperative CT scans because the associated osseous abnormalities, which included slight narrowing of the MF and little cortical irregularity, were not always appreciated ( 20 ). In another study ( 11 ), both CT and magnetic resonance (MR) imaging were used to evaluate patients clinically suspected of having coalition. Of four patients with surgically confirmed subtalar coalitions, two had normal preoperative CT images, whereas all four had abnormal MR images. The two cases that were missed at CT were both fibrous coalitions histologically, adding support to the view that CT may be limited in its depiction of fibrous subtalar coalitions ( 8,20 ). Linklater and associates ( 8 ) also noted that fibrous extraarticular coalitions occurring posterior to the MF were associated with variable, often subtle, bone abnormalities in the adult patient that were thought to be secondary to abnormal forces transmitted through the coalition. Radiology: Volume 257: Number 3 December 2010 n radiology.rsna.org 833

5 Table 1 Figure 5 Comparison of Groups with Respect to Age, Sex, and CT Section Thickness Variable Coalition Group ( n = 35) Control Group ( n = 33) P Value Age (y) Male sex * 21 (60) 21 (64).758 CT section thickness (mm) Note. Unless otherwise indicated, data are means 6 standard deviations. * Data are numbers of patients, with percentages in parentheses. Table 2 Comparison of Measurements between Groups Variable Coalition Group Control Group P Value * MF (mm) ,.001 ST (mm) MF/ST ratio ,.001 ST posterior to MF (mm) ,.001 Note. Data are means 6 standard deviations. * P values derived by using Wald tests for group effects adjusted for age, sex, and CT section thickness in a linear regression model. Our observations reveal alterations in the morphology of the ST and MF in the contralateral unaffected ankle in pediatric patients with unilateral subtalar coalition. We found that the ST was larger and extended more posterior to the MF than in control subjects, and when the ST was elongated it was associated with a hypoplastic MF. These findings suggest that some patients with subtalar coalition have bilateral developmental abnormalities of the talus and calcaneus that clinically may be expressed initially in only the more severely affected ankle. We have been unable to find any prior reports detailing the anatomic variations in the ST in the contralateral ankle in patients with unilateral subtalar coalition. Given that tarsal coalitions are thought to arise secondary to a failure of mesenchymal differentiation and segmentation during fetal development ( 1 ), it is not surprising that a morphologic abnormality may exist on both sides when only a unilateral coalition is detected. Wechsler et al ( 11 ) and Kumar et al ( 21 ) showed that some patients have symptomatic nonosseous coalitions that are not evident at CT. A morphologic alteration in the ST might provide a clue to the lesion. Because subtalar coalition is often a bilateral disease (bilateral subtalar coalitions were found in 46 of 91 patients in our study), some patients with unilateral coalition may have a morphologic abnormality of the contralateral ankle that reflects the mildest expression of bilateral disease in the absence of synchronous contralateral symptoms or other CT findings. Future prospective studies will be important in determining whether these anatomic observations may be related to subtle, fibrous coalitions along the posterior side of the ST or whether patients with a diagnosis of unilateral subtalar coalition later develop symptoms on the contralateral side. This study had several limitations. The control population consisted of patients with triplane fractures. Although these patients may have had preexisting symptoms, it is a reasonable assumption that these patients otherwise had no symptoms. Given the retrospective study design, the study patients were evaluated by various clinicians before imaging. It is possible that specific symptoms were not noted in the medical record. The mean age of the patients in the study population was slightly older than that of the patients in the control population ( Table 1 ), though we took age into account as part Figure 5: (a) Three-dimensional reconstructed CT image obtained by using a bone algorithm in 13-year-old adolescent boy with a contralateral nonosseous subtalar coalition demonstrates posterior extension of the ST without frank coalition (arrow). (b) Three-dimensional CT reconstruction in 11-year-old boy with triplane fracture demonstrates a normal ST (arrow). of the linear regression model used for statistical analysis. Although the CT scanning protocol was not uniform across the course of the study, there was no significant difference in CT section thickness between groups ( Table 1 ). In addition, the CT section thickness was adjusted for in the statistical analysis comparing groups. Some patients diagnoses were apparent on the original CT scans, which may have introduced reader bias. We limited this bias by having readers obtain measurements only on the reformatted images where the diagnosis was most often excluded from the field of view. An inherent measurement error was associated with the placement of the electronic calipers, especially in patients with a dysplastic ST. The near-perfect intraobserver variability in measurements ( 22 ) 834 radiology.rsna.org n Radiology: Volume 257: Number 3 December 2010

6 suggests that measurement error was not a significant limitation. Importantly, this cross-sectional study could not be used to determine if any of the patients with unilateral coalition and morphologic abnormalities of the contralateral ankle developed symptomatic bilateral disease. In conclusion, our findings support the view that subtalar coalition is not a lesion localized to the MF but rather reflects a diffuse process that can extend along the greater length of the talocalcaneal interval. This result points to a fundamental morphologic alteration of the subtalar articulation that could provide insight into the etiology of this important orthopedic disorder. The presence of this finding at CT may have implications regarding the asymptomatic ankle in patients with a contralateral coalition, as well as for patients with clinical concerns of coalition but otherwise normal CT findings, and longitudinal studies in this regard are encouraged. References 1. Harris RI. Rigid valgus foot due to talocalcaneal bridge. J Bone Joint Surg Am 1955 ; 37-A ( 1 ): Anton C. Tarsal coalition. In: Donnelly LF, ed. Diagnostic imaging: pediatrics. Salt Lake City, Utah : Amirsys, 2005 ; El Rassi G, Riddle EC, Kumar SJ. Arthrofibrosis involving the middle facet of the talocalcaneal joint in children and adolescents. J Bone Joint Surg Am 2005 ; 87 ( 10 ): Stormont DM, Peterson HA. The relative incidence of tarsal coalition. Clin Orthop Relat Res 1983 ;( 181 ): Rühli FJ, Solomon LB, Henneberg M. High prevalence of tarsal coalitions and tarsal joint variants in a recent cadaver sample and its possible significance. Clin Anat 2003 ; 16 ( 5 ): Solomon LB, Rühli FJ, Taylor J, Ferris L, Pope R, Henneberg M. A dissection and computer tomograph study of tarsal coalitions in 100 cadaver feet. J Orthop Res 2003 ; 21 ( 2 ): Vincent KA. Tarsal coalition and painful flatfoot. J Am Acad Orthop Surg 1998 ; 6 ( 5 ): Linklater J, Hayter CL, Vu D, Tse K. Anatomy of the subtalar joint and imaging of talocalcaneal coalition. Skeletal Radiol 2009 ; 38 ( 5 ): Nalaboff KM, Schweitzer ME. MRI of tarsal coalition: frequency, distribution, and innovative signs. Bull NYU Hosp Jt Dis 2008 ; 66 ( 1 ): Crim JR, Kjeldsberg KM. Radiographic diagnosis of tarsal coalition. AJR Am J Roentgenol 2004 ; 182 ( 2 ): Wechsler RJ, Schweitzer ME, Deely DM, Horn BD, Pizzutillo PD. Tarsal coalition: depiction and characterization with CT and MR imaging. Radiology 1994 ; 193 ( 2 ): Wechsler RJ, Karasick D, Schweitzer ME. Computed tomography of talocalcaneal coalition: imaging techniques. Skeletal Radiol 1992 ; 21 ( 6 ): Westberry DE, Davids JR, Oros W. Surgical management of symptomatic talocalcaneal coalitions by resection of the sustentaculum tali. J Pediatr Orthop 2003 ; 23 ( 4 ): Laor T. Congenital malformations of bone. In: Slovis TL, ed. Caffey s pediatric diagnostic imaging. 11th ed. Philadelphia, Pa : Mosby, 2008 ; Chew F. Developmental and congenital conditions. In: Musculoskeletal imaging. Philadelphia, Pa : Lippincott Williams & Wilkins, 2003 ; Newman JS, Newberg AH. Congenital tarsal coalition: multimodality evaluation with emphasis on CT and MR imaging. RadioGraphics 2000 ; 20 ( 2 ): Emery KH, Bisset GS 3rd, Johnson ND, Nunan PJ. Tarsal coalition: a blinded comparison of MRI and CT. Pediatr Radiol 1998 ; 28 ( 8 ): Lateur LM, Van Hoe LR, Van Ghillewe KV, Gryspeerdt SS, Baert AL, Dereymaeker GE. Subtalar coalition: diagnosis with the C sign on lateral radiographs of the ankle. Radiology 1994 ; 193 ( 3 ): Resnick D. Talar ridges, osteophytes, and beaks: a radiologic commentary. Radiology 1984 ; 151 ( 2 ): Lee MS, Harcke HT, Kumar SJ, Bassett GS. Subtalar joint coalition in children: new observations. Radiology 1989 ; 172 ( 3 ): Kumar SJ, Guille JT, Lee MS, Couto JC. Osseous and non-osseous coalition of the middle facet of the talocalcaneal joint. J Bone Joint Surg Am 1992 ; 74 ( 4 ): Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977 ; 33 ( 1 ): Radiology: Volume 257: Number 3 December 2010 n radiology.rsna.org 835