| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Walter Reed Army Medical Center, 6900 Georgia Ave.,
NW, Washington, DC 20307-5001.
2 Department of Radiology and Nuclear Medicine, Uniformed Services, University
of the Health Sciences, 4301 Jones Bridge Rd., Bethesda, MD 20814.
3 Department of Orthopedic Surgery, Walter Reed Army Medical Center, Washington,
DC 20307-5001.
4 Department of Radiologic Pathology, Armed Forces Institute of Pathology,
Washington, DC 20306-6000.
Received March 3, 2004;
accepted after revision May 21, 2004.
Address correspondence to P. A. Dinauer
(phil.dinauer{at}us.army.mil).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We retrospectively collected 58 hip MRI studies along with surgical reports in 58 patients who underwent hip arthroscopy over a 5-year period. Intraoperative photography (n = 23), radiography (n = 56), unenhanced MRI (n = 13), and MR arthrography (n = 54) studies were available for review. Two radiologists described hip anatomy on radiology studies with agreement by consensus.
RESULTS. A normal posteroinferior sublabral groove was confirmed on available arthroscopy photographs in four (17.4%) of 23 hips. In each of these four patients, the anatomic sublabral groove correlated with apparent partial labral detachment on MR arthrography. On review of all studies, 13 hips (22.4%) without a posterior labral tear at surgery had imaging findings of a sublabral sulcus. The sulcus was not associated with acetabular dysplasia, which was radiographically noted in 12 cases (21.4%). Preoperatively, the sulcus was misdiagnosed as a tear in two cases. Labral tears were anterior or anterosuperior in 51 patients.
CONCLUSION. A posteroinferior sublabral groove is a relatively common normal anatomic hip variation. If not recognized as normal, the sulcus may serve as a diagnostic pitfall on MR arthrography. Its location is distinct from most labral tears. We did not discover a sublabral sulcus at the anterior or anterosuperior acetabulum, the most common sites of labral injury.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MR arthrography, typically performed after injection of dilute gadolinium or normal saline into the hip joint [5, 6], allows accurate characterization of intrasubstance labral tears and detachments [6, 15]. The normal acetabular labrum is triangular in shape and usually is hypointense on MR images [16], particularly in patients younger than 50 years [17]. With MR arthrography, labral tears are detected when injected fluid extends into the articular surface of the labrum. Fluid undermines the labral base in labral detachments. However, several diagnostic imaging pitfalls have been recognized. Signal alterations at the junction of acetabular hyaline cartilage with labral fibrocartilage may be mistaken for a labral detachment [2, 16]. Variations in labral signal intensity, particularly with increasing age, have been documented on unenhanced MRI examinations of asymptomatic hips [17–19]. Also, there has been speculation that either a congenital or degenerative absence of a portion of the labrum may serve as a diagnostic pitfall [17, 18].
Perhaps the most elusive and controversial potential pitfall is that of a sublabral sulcus. The diagnostic pitfall of a single case of a normal sublabral sulcus has been presented with operative and MR arthrography correlation (Marianacci EB et al., presented at the 1995 annual meeting of the Radiological Society of North America). Other investigators have suggested that a normal sublabral sulcus may exist at the anterosuperior aspect of the acetabulum [7, 20, 21]. In 1996, Byrd [7] recorded his surgical observation that patients with mild acetabular dysplasia may have a normal partial separation of the labrum from the superior acetabular rim (i.e., a normal sublabral sulcus). However, a normal sublabral foramen or sulcus has not been confirmed in other surgical or cadaveric studies of adult hips [2, 5, 10, 16].
In 1994, Keene and Villar [22] presented normal anatomic findings at hip arthroscopy, including a normal sublabral groove at the inferior aspect of the acetabulum, adjacent to the acetabular fossa. This sublabral groove is a potential pitfall in diagnosing labral tears on MR arthrography [23]. However, after conducting a search of the literature, we have been unable to find published examples of surgical and imaging correlation for this normal anatomic variant. At our institution, patients with mechanical hip pain and clinically suspected acetabular labral tears routinely undergo MR arthrography evaluation of their symptomatic hip before arthroscopic surgery. The primary purpose of our retrospective study was to correlate findings on hip MRI and MR arthrography with arthroscopic surgery to assess the location, prevalence, and potential diagnostic pitfalls of a normal acetabular sublabral sulcus.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The study population consisted of 27 women and 31 men with an average age of 35 years (range, 16–56 years). Radiographs of the affected hip were available in 56 patients. MR arthrography studies of 54 hips were obtained. Nine of these 54 patients also had a prior conventional MRI study of their affected hip. A conventional unenhanced MRI was the only preoperative imaging study performed for four additional patients. Hip arthroscopy photographs of 23 patients were collected for retrospective review.
The MR arthrography studies were obtained using either saline (n = 13) or gadolinium (n = 41) as the primary contrast agent on chemical shift selective fat-suppressed T1-weighted or fat-suppressed fast spin-echo T2-weighted images. Arthrography procedures were performed under fluoroscopic guidance immediately before each MR arthrography study. A sterile arthrography tray was prepared for each patient. Under fluoroscopic guidance, a 22-gauge needle was advanced onto the anterior femoral neck, lateral to the femoral vessels. For gadolinium studies, an intraarticular volume of 10–17 mL was injected from a syringe containing a mixture of 4 mL of iodinated contrast material ([iothalamic acid] Conray-60, Mallinckrodt), 10 mL of 1% lidocaine, and 0.05 mmol of dilute gadolinium ([gadopentetate dimeglumine] Magnevist, Berlex), resulting in a gadolinium dose of 0.003 mmol/mL. For saline arthrography, a mixture of 6 mL of Conray-60 with 10 mL of 1% lidocaine was prepared. In saline and gadolinium arthrography, the use of iodinated contrast material allowed confirmation of proper intraarticular fluid placement before sending the patient to the MRI suite.
All MRI studies were performed on a closed 1.5-T unit (Signa, GE Healthcare). A torso phased-array coil was placed over the affected hip. The imaging protocol for gadolinium arthrography consisted of fat-suppressed T1-weighted axial, coronal oblique, and sagittal oblique images (TR range/TE range, 450–850/9–15) at a 16- to 20-cm field of view, a 3- to 4-mm slice thickness, and a gap of 0–1 mm. The sagittal and coronal oblique images were acquired perpendicular to the orientation of the acetabulum, as described by Czerny et al. [6]. The imaging matrix was 192 (or 224) x 256 with 2–4 acquisitions. For saline arthrography, fat-suppressed fast spin-echo T2-weighted images were acquired in all three imaging planes using a 16- to 18-cm field of view. On fast spin-echo images, the TR ranged from 1,750 to 4,000, TE was 35–90, and echo-train length was 8–12. The matrix and slice thickness were the same as described for gadolinium arthrography. On conventional MRI studies (n = 13), the hip was evaluated in three standard planes, the field of view was variable (16–24 cm), and slice thickness was 3–4 mm on T1- and fat-suppressed fast spin-echo T2-weighted images.
The radiologists reviewed surgical reports and available arthroscopy photographs with the orthopedic surgeon before reanalyzing the MRI studies. In standard hip arthroscopy procedures, the surgeon places the patient in a supine position on a fracture table. While traction is applied to the hip, anterior and lateral portals are created to allow arthroscopic inspection of the anterior, superior, and posterior portions of the hip joint. Sites of labral tearing and detachment are recorded on the basis of direct inspection of the articular surface of the labrum and use of a probing instrument. Labral tears and avulsions show abnormal displacement in response to probing (Fig. 1A, 1B). Arthroscopically, a sublabral groove was defined as a smooth, well-defined cleft separating the hyaline cartilage from the labrum. This groove had no accompanying evidence of tissue trauma, and it did not produce an abnormal, unstable labral base in response to probing. Although the operative reports did not routinely describe the presence or absence of a normal sublabral groove, the posterior groove, when identified by the surgeon, was documented on intraoperative photographs.
|
|
Two radiologists with knowledge of the surgical findings retrospectively reviewed the 54 MR arthrography and 13 conventional MRI studies. The radiologists reached agreement by consensus in describing labral anatomy and pathology as depicted on the imaging studies. A sublabral sulcus was considered definitively present when MRI findings of a sulcus correlated directly with a normal sublabral groove as depicted on arthroscopy photographs. Also, in the 35 cases without available intraoperative photographs, the reviewers studied the anterior, superior, and posterior portions of the acetabular labrum for sites of a presumed normal sublabral sulcus. A sulcus was considered likely present when labral separation from the articular cartilage was clearly observed on at least two MR images in the absence of surgical pathology in the same region of the acetabulum.
Anterior–posterior hip radiographs were reviewed by the two radiologists who reached consensus in assessing acetabular shape and adequacy of femoral head coverage. Acetabular dysplasia was considered present when the center-edge angle of Wiberg measured less than 25°, the horizontal toite externe angle measured more than 10°, or at least 25% of the femoral head was uncovered by the shallow acetabulum [24].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Anterior–posterior radiographs were available for retrospective review in 56 patients. Twelve (21.4%) of the 56 had acetabular dysplasia. Nine hips were dysplastic as manifested by a center-edge angle of less than 25° along with a horizontal toite externe angle of more than 10°. Three additional patients had an abnormal horizontal toite externe angle (> 10°), but a normal center-edge angle. Ten of the 12 dysplastic hips also had femoral head uncovering of at least 25%. The other 44 patients with available radiographs had normal acetabular shape and adequate femoral head coverage.
A well-defined, normal sublabral groove at the posteroinferior acetabulum was identified in four of the 23 patients with intraoperative photographs available for retrospective review (Figs. 2A, 2B and 3A, 3B). The arthroscopist did not observe a sublabral groove or sulcus as a normal variation in any locations other than at the posteroinferior acetabulum. Surgically, a sublabral groove was characterized by a smoothly marginated cleft at the labral base. This groove did not produce an unstable labrum in response to direct probing. In each of the patients with a sublabral groove at arthroscopy, MR arthrography showed partial separation of the posteroinferior labrum from the underlying articular cartilage on at least two images. None of the patients with a confirmed sublabral groove underwent conventional MRI study before surgery. Therefore, the visibility of this normal variant on unenhanced images could not be evaluated. On the MR arthrography examinations displaying a sublabral sulcus, the posterior labrum was otherwise normal with low signal intensity and normal triangular morphology.
|
|
|
|
On retrospective review of all available intraoperative photographs, we could find no cases in which MR arthrography falsely depicted a sublabral groove at the posterior acetabulum. In the 35 hips without intraoperative photographs available for retrospective review, a sublabral sulcus or groove at the posterior acetabulum was identified in nine patients. None of these patients had a traumatic posterior labral tear or detachment described at surgery. Therefore, the prevalence of the posterior sublabral sulcus based on review of all available MRI studies was 13 (22.4%) of 58 patients. In two patients with a prominent posterior sublabral sulcus that was well visualized on both axial and sagittal images, the preoperative MR arthrography reports described a possible posterior labral tear.
On retrospective review of all MRI examinations, three (5.2%) of 58 hips had labral tears in which MR images showed an apparent absent portion of the labrum on at least one slice (Fig. 4). These patients had tears that could account for the apparent labral absence on a single MR image: a severe degenerative tear (n = 1), a large radial tear (n = 1), and a displaced labral flap (n = 1). None of the 58 patients had a documented congenital absence of a portion of the labrum. A normal cleft was observed at the junction of the anterior labrum with the transverse ligament in 19 (32.8%) of 58 hips. In one case from the author's early experience with hip MR arthrography, an anterior cleft was mistaken for a labral tear. The cleft at the anterior labral–ligamentous junction was frequently accompanied by a periligamentous recess located anterior to the ligamentum teres and medial to the anterior portion of the transverse ligament (Fig. 5A, 5B). This recess was typically best seen on coronal images.
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several recent articles provide an excellent review of normal anatomy as depicted on hip MR arthrography [14, 16, 21]. The normal acetabular labrum, composed of fibrocartilage [2], is typically triangular in shape and is frequently homogeneously hypointense on MR images [16, 19] (Fig. 6A, 6B, 6C). However, labral signal intensity can increase with age [17, 19] or may be increased at its base in the absence of degeneration [2], resulting in a potential imaging pitfall of tear overdiagnosis. The labrum firmly adheres to the acetabular rim except at the margins of the acetabular fossa where it can be separated from the underlying hyaline cartilage by a distinct groove [22]. Inferiorly, the labrum is continuous with the transverse ligament that bridges the acetabular notch. Generally, in our evaluation of patients with mechanical hip pain, the imaging finding of labral separation from the underlying acetabulum by joint fluid or injected contrast material is considered a tear until proven otherwise.
|
|
|
Unlike the shoulder, where a normal sublabral foramen has been identified at the anterosuperior quadrant of the glenoid [29, 30], a normal sublabral foramen, or sublabral sulcus, has not been convincingly shown in the hip [5, 16]. Several investigators have suggested that a normal sublabral sulcus, mimicking a traumatic labral detachment, may exist at the anterosuperior acetabulum [7, 14, 20]. In our opinion, normal variations in labral shape, such as the round labral shape described by Lecouvet et al. [19], may give the false impression of a sublabral sulcus on MR arthrography. In 1996, Byrd [7] described a normal anatomic variation of partial separation of the labrum from the superior acetabulum in patients with acetabular dysplasia. Presumably, patients with acetabular dysplasia can develop a sublabral sulcus in response to chronic increased stress loads applied to the superior labrum from inadequate anterosuperior femoral head coverage [31]. Oneda et al. [32] observed increased stress distribution in the anterosuperior aspect of the hip at walking, and therefore this region of the acetabulum would seem predisposed to develop either what has been interpreted as a "normal anomalous variation" [7] or a labral tear requiring excision and reconstructive surgery [1]. In our experience with MR arthrography, the finding of injected contrast material that partially or completely undermines the anterosuperior labral base has surgically correlated with labral injury (Fig. 7A, 7B, 7C).
|
|
|
A normal sublabral sulcus has not been shown in several studies involving cadaveric or surgical correlation [2, 5, 10, 16]. In a study of 170 hips with mild or moderate acetabular dysplasia, McCarthy and Lee [33] discovered labral tears in 72%. These authors did not describe a normal anterior or superior sublabral sulcus as a diagnostic pitfall. Also, a superior sublabral sulcus was not included as a normal anatomic variant in the 1994 arthroscopy article by Keene and Villar [22]. In addition, none of the 58 patients in our series, including the 12 with acetabular dysplasia, had MRI or surgical findings confirming the presence of a normal sublabral sulcus at the anterosuperior acetabular rim.
Several studies that discuss normal labral variations, such as a sublabral sulcus or absence of the labrum, are limited by lack of surgical or direct anatomic correlation [17–19]. In unenhanced MRI studies of 71 asymptomatic hips, Abe et al. [17] found absence of a portion of the labrum in some patients 50 years old and older. In a study describing labral variations in 200 asymptomatic hips, Lecouvet et al. [19] only obtained unenhanced coronal T1-weighted images; their finding of an increased incidence of absent labra with increasing age may have reflected increased labral signal from labral degeneration rather than a true labral absence. In another study involving unenhanced MRI evaluation of 52 asymptomatic hips, the authors observed that 10% of patients had an absent portion of the anterosuperior labrum [18]. This study did not include sagittal MR images that, in our experience, usually best show anterior or anterosuperior labral injuries. In our study group, three of 58 patients (5.2%) had labral tears in which MR images showed an apparent absent portion of the labrum. In these cases, the anterosuperior labrum was severely frayed, had a radial tear, or had a detached portion. We found no examples of congenital or developmental absence of the acetabular labrum. Therefore, in our opinion, it is difficult to draw solid conclusions about normal labral variations from prior research studies performed in limited imaging planes without contrast enhancement and without direct anatomic correlation.
In a report of normal arthroscopic anatomy of the hip based on a review of 100 cases, Keene and Villar [22] described a normal variant at the margin of the acetabular fossa. Their 1994 article showed two examples of a normal sublabral groove located at the posteroinferior acetabulum. This groove represents a well-defined, stable separation between the labrum and articular cartilage. Although this groove has been recognized arthroscopically, it has not been described in a large group of patients with both surgical and imaging correlation. The normal groove is important to recognize because it has been described as a potential pitfall in imaging diagnosis of labral tears [23], but we have been unable to find published MR images of this normal variant. The unique posteroinferior location of the sublabral sulcus is helpful in differentiating it from most labral injuries, which are anterior or anterosuperior [3, 11, 16].
Results from our study are in agreement with the arthroscopic observations made by Keene and Villar [22]. Specifically, in our case series, we were unable to identify a normal sublabral sulcus in any locations other than at the posteroinferior margin of the acetabular fossa. The posterior sublabral sulcus was not mentioned in most preoperative MRI reports given the presence of a clearly visualized labral tear or detachment at the more anterior aspect of the acetabulum and given the smooth contour of what was suspected to be an insignificant, normal posterior labral variant. However, in two patients from our series, the MRI reports indicated concern for the presence of a posterior labral tear. The diagnostic pitfall encountered in these two cases reflects the occasional prominent nature of the sulcus, which was well seen on multiple fat-suppressed axial and sagittal-oblique images. Each of these patients had an anterior labral tear and a normal posterior labrum with sublabral groove as shown on arthroscopy photographs. It is unclear why a posterior sulcus was not noted in the 40 patients and six cadavers studied by Czerny et al. [16], but their patients' hips were evaluated by arthrotomy rather than arthroscopy, and they used a gradient-echo sequence that differs from our imaging technique.
The posteroinferior sublabral groove, retrospectively identified in 22.4% of our 58 patients, was not associated with acetabular dysplasia. Only one of 12 hips with acetabular dysplasia had a sublabral groove on MR arthrography. The location of this relatively common normal variant is diagnostically advantageous because only a small percentage of labral tears are confined to the posterior acetabulum. In 2001, McCarthy et al. [11] reported that 11% of 261 labral tears in their series were posterior in location. In our study population, no patients had isolated posterior labral tears, and of the 54 patients with labral tears, 5.6% had combined anterior and posterior labral injury. However, in the experience of some researchers in Japan, distribution of labral lesions has been more posterior than anterior [26, 28, 34]. The pathogenesis of these posterior injuries in Japanese patients is unclear, but may be related to increased, repetitive posterior hip stress arising from prolonged, deep hip flexion from a cultural habit of squatting or sitting on the floor [26].
In addition to the posterior sublabral sulcus, another normal variation should be recognized on MR arthrography examinations. A cleft formed at the junction of the transverse ligament with the labrum may give the false impression of a labral tear [4, 21]. In our series, 32.8% of hips showed this cleft at the anterior labral–ligamentous junction. This normal finding is located more inferior than most anterior labral tears and should not be misinterpreted as a traumatic anteroinferior labral–ligamentous detachment. The cleft often was accompanied by an adjacent periligamentous recess located anteromedial to the ligamentum teres.
Our study is partially limited by its retrospective nature. A list of normal arthroscopic findings was inconsistently provided in the surgical reports, which tended to focus on technical details and abnormal observations. A well-designed prospective study in which MRI findings are correlated with arthroscopic findings recorded in all portions of the labrum would further advance our understanding about the incidence and location of normal labral variations. In our study, intraoperative photographs were available in 23 of 58 patients. These photographs, as well as intraoperative videotapes, are valuable educational supplements to the operative reports. Nearly all patients studied had anterior or anterosuperior labral tears. Therefore, we question whether these tears could obscure a preexisting sublabral sulcus. However, none of the 23 patients with intraoperative photographs available for review had definite surgical evidence of both a sublabral sulcus and labral tear in the same region of the acetabulum.
Another study limitation is population bias. Patients selected for arthroscopic surgery had hip pain or mechanical symptoms of labral tear, and most had labral tears defined on MR images. Therefore, a population bias favoring surgical confirmation of labral injury was introduced. In our study population of 58 patients with an average age of 35 years, 93.1% had surgically confirmed labral tears. No asymptomatic hips received radiologic and arthroscopic evaluation. Because our retrospective study lacks an age-matched control group, it is unclear how frequently contrast-enhanced MR images would depict labral tears in asymptomatic hips. On the basis of the study by Abe et al. [17], we would expect fewer than 10% of asymptomatic hips in subjects less than 40 years old to have imaging findings of labral tearing or severe labral degeneration. The natural history and clinical significance of asymptomatic labral injuries are unknown; the incidental MRI finding of an acetabular labral tear may not warrant surgical evaluation and treatment. However, acetabular labral injury frequently is associated with articular cartilage damage [11, 12], possibly related to femoroacetabular impingement disorder or altered hip biomechanics, and labral tears may contribute to the development of osteoarthritis [11, 12, 35].
In conclusion, MR arthrography of the hip plays an important role in defining labral pathology, excluding extraarticular sources of pain, and helping the orthopedic surgeon decide which patients may benefit from hip arthroscopy. Consequently, radiologists must be familiar with normal anatomic variations to avoid potential pitfalls in imaging diagnosis of labral tears. The posteroinferior sublabral groove, or sulcus, is a relatively common normal variation that, to our knowledge, has not been studied previously in a large group of patients with both MRI and arthroscopy correlation. This normal sublabral sulcus is not associated with acetabular dysplasia. In some cases, the sulcus can be quite prominent, resulting in the potential pitfall of a false-positive diagnosis of posterior labral tear or detachment. In our experience, the sulcus is confined to the posteroinferior acetabulum, whereas most labral tears are located in the anterior or anterosuperior portions of the acetabulum. We were unable to confirm the presence of other suspected normal labral variations, such as an anterosuperior sublabral sulcus or a congenital absence of a labral segment, either because they do not exist or are very rare. This information should improve diagnostic accuracy when interpreting hip MR arthrography.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Cashin, H. Uhthoff, M. O'Neill, and P. E. Beaule Embryology of the acetabular labral-chondral complex J Bone Joint Surg Br, August 1, 2008; 90-B(8): 1019 - 1024. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. J. James, K. Ali, F. Malara, D. Young, J. O'Donnell, and D. A. Connell MRI findings of femoroacetabular impingement. Am. J. Roentgenol., December 1, 2006; 187(6): 1412 - 1419. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Saddik, J. Troupis, P. Tirman, J. O'Donnell, and R. Howells Prevalence and location of acetabular sublabral sulci at hip arthroscopy with retrospective MRI review. Am. J. Roentgenol., November 1, 2006; 187(5): W507 - W511. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L.-C. Lien, J. C. Hunter, and Y.-S. Chan Tubular acetabular intraosseous contrast tracking in MR arthrography of the hip: prevalence, clinical significance, and mechanisms of development. Am. J. Roentgenol., September 1, 2006; 187(3): 807 - 810. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. A. Toomayan, W. R. Holman, N. M. Major, S. M. Kozlowicz, and T. P. Vail Sensitivity of MR Arthrography in the Evaluation of Acetabular Labral Tears Am. J. Roentgenol., February 1, 2006; 186(2): 449 - 453. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
