MR Arthrography of the Hip: Diagnostic Performance of a Dedicated ...

17 downloads 4205 Views 2MB Size Report
dedicated cartilage MR sequence (water-excitation 3D double-echo steady-state) with a stan- dard MR sequence (T1-weighted spin-echo) in detecting articular ...
MR Arthrography of the Hip: Diagnostic Performance of a Dedicated Water-Excitation 3D Double-Echo Steady-State Sequence to Detect Cartilage Lesions Patrick R. Knuesel1,2 Christian W. A. Pfirrmann1 Hubert P. Noetzli3,4 Claudio Dora3 Marco Zanetti1 Juerg Hodler1 Bernd Kuehn5 Marius R. Schmid1

OBJECTIVE. The objective of our study was to compare the diagnostic performance of a dedicated cartilage MR sequence (water-excitation 3D double-echo steady-state) with a standard MR sequence (T1-weighted spin-echo) in detecting articular cartilage lesions of the hip after intraarticular injection of gadopentetate dimeglumine. MATERIALS AND METHODS. In 50 MR arthrograms of the hip joint obtained in 47 consecutive patients, a sagittal 3D double-echo steady-state sequence (TR/TE, 24/6.5; flip angle, 25°) was compared with a sagittal T1-weighted spin-echo sequence (350/14). Two musculoskeletal radiologists independently evaluated articular cartilage. Sensitivity and specificity for detecting cartilage defects were calculated for those hips that underwent open surgery (n = 21). Lesion conspicuity was retrospectively reviewed and graded between 1 (not visible) and 5 (well defined). RESULTS. At surgery, a total of 26 lesions of the acetabular (n = 20) and femoral (n = 6) cartilage were found. For the 3D double-echo steady-state and T1-weighted spin-echo sequences, sensitivities and specificities for cartilage lesion detection were 58% and 88% and 81% and 81% for reviewer 1 and 62% and 94% and 62% and 100% for reviewer 2, respectively. Lesion conspicuity was significantly superior (p = 0.036) for the 3D double-echo steady-state sequence (mean grade, 3.4) compared with the T1-weighted spin-echo sequence (mean grade, 3.0). The kappa value was fair for the 3D double-echo steady-state sequence (κ = 0.40) and moderate for the T1-weighted spin-echo sequence (κ = 0.55). CONCLUSION. The 3D double-echo steady-state sequence optimized for cartilage imaging improves lesion conspicuity but does not improve diagnostic performance.

O

Received November 25, 2003; accepted after revision May 20, 2004. 1 Department of Radiology, University Hospital Balgrist, Forchstrasse 340, Zurich CH-8008, Switzerland. Address correspondence to M. R. Schmid ([email protected]). 2 Present address: Department of Radiology, Kantonsspital Baden, Daettwil CH-5405, Switzerland. 3

Department of Orthopedics, University Hospital Balgrist, Zurich CH-8008, Switzerland.

4 Present address: Department of Orthopedic Surgery, Zieglerspital Bern, Morillonstrasse, Bern 3001, Switzerland. 5MR Application Development, Siemens AG Medical Solutions, Karl Schall Strasse 4, Erlangen D-91050, Germany.

AJR 2004;183:1729–1735 0361–803X/04/1836–1729 © American Roentgen Ray Society

AJR:183, December 2004

steoarthritis of the hip is not limited to elderly patients but may also be found in middle-aged adults and even young adults. In such patients, surgical dislocation of the hip with shaving of cartilage irregularities and surgical correction of any deformity of the femoral head–neck junction may be considered [1]. This type of surgery may result in substantial pain relief, restore function, and postpone prosthetic treatment [1–4]. For this recently introduced surgical technique, early detection of cartilage damage is relevant for proper patient selection. Standard radiographs are not adequate for this purpose because radiographic signs such as osteophytes, subchondral cysts, and subchondral sclerosis may not correlate with the degree of articular cartilage damage [5]. MRI with and without intraarticular administration of gadopentetate dimeglumine has been widely used for the assessment of

articular cartilage [5–31]. Contrary to many investigations reporting excellent results in imaging cartilage of the knee joint [6–10], evaluation of articular cartilage in the hip joint is more difficult [5, 11]. This difference may be explained by the fact that cartilage of the hip joint is thinner than that in the knee joint and larger coils with lower signal-tonoise ratio are used. Therefore, optimization of MR sequences for cartilage imaging in the hip is relevant. Several MR sequences have been used for cartilage imaging, usually in the knee joint, including fat-suppressed 3D spoiled gradient-recalled acquisition in the steady state (SPGR) [7, 12–15, 23], fast low-angle shot (FLASH) [8, 18], T2- and intermediateweighted fat-suppressed fast spin-echo [6, 9], and water-excitation 3D double-echo steady-state [10, 16] sequences. In the detection of cartilage defects in cadaveric knee

1729

Knuesel et al. joints, MR arthrography has been shown to be superior to MRI without intraarticular contrast administration [14]. A combination of MR sequences specific for cartilage and intraarticular gadolinium injection could probably improve imaging of the cartilage in the hip. The 3D double-echo steady-state sequence allows thinner sections to be obtained than conventional spin-echo sequences and provides high contrast between intraarticular structures and injected gadopentetate dimeglumine. The purpose of this study was to compare the diagnostic performance of an MR sequence dedicated to cartilage imaging (3D double-echo steady-state) with a standard MR sequence (T1-weighted spin-echo) in detecting articular cartilage lesions of the hip after intraarticular injection of gadopentetate dimeglumine. Materials and Methods Patients and Standard of Reference All hip MR arthrograms obtained between March 2002 and December 2002 of patients with clinically suspected articular cartilage damage or suspected labral lesions were retrospectively analyzed. Fifty consecutive examinations (27 right hips and 23 left hips) in 47 patients (23 women, 24 men; mean age, 38.5 years; age range, 20–73

years) were identified. A subgroup comprising 21 of the 50 hips underwent open surgery with hip dislocation according to the technique described by Ganz et al. [2]. This technique allows complete inspection of both the femoral and acetabular cartilage. All surgical procedures were performed by the same two hip surgeons. Surgical reports are standardized in our institution and precisely describe the extent and location of all cartilage lesions. The time interval between MR arthrography and surgery varied between 2 and 6 months. The hospital’s review board allowed us to perform this type of study on the basis of a general permit issued by the responsible state agency. Imaging Protocol All studies were performed on a 1.5-T MR system (Symphony, Siemens Medical Solutions) according to a standardized protocol. A flexible wraparound receive-only surface coil was used. A sagittal 3D double-echo steady-state sequence (Fig. 1) optimized for cartilage imaging and a sagittal T1-weighted spin-echo sequence were performed. In this 3D double-echo steady-state sequence, a 1–2–1 binomial pulse is used for spatial- and frequency-selective excitation of water isochromats while suppressing signals from fatty tissues. The double-echo steady-state sequence simultaneously acquires the signals of two sequences. The first one is the signal of a fast imaging with steady-state precession (FISP) sequence, also known as coherent steady-state sig-

Fig. 1.—Schematic drawing illustrates 3D double-echo steady-state sequence. Double-echo steady-state sequence simultaneously acquires signals of two sequences. First one is signal of fast imaging with steady-state precession (FISP) sequence, also known as coherent steady-state signal. Second one is signal of PSIF sequence (time reversal of FISP). RF = radiofrequency, GS = slice-selection gradient, GP = phase-encoding gradient, GR = readout gradient, ADC = analog-to-digital converter.

1730

nal. The second one is the signal of a PSIF sequence (time reversal of FISP). Images from both echoes are reconstructed and added using a sum-ofsquares algorithm. Adding the FISP and PSIF signals increases the T2-weighting of the FISP images and allows improved delineation of fat, joint fluid, and cartilage in comparison with a FISP image. On water-excitation 3D double-echo steady-state images, both joint fluid and gadopentetate dimeglumine (2 mmol/L) are hyperintense, whereas articular cartilage has an intermediate signal intensity. The imaging parameters for the 3D doubleecho steady-state sequence in our study were as follows: TR/TE, 24/6.5; section thickness, 1.7 mm; no intersection gap; flip angle, 25°; field of view, 15 cm; matrix, 512 × 512; 1 average; and acquisition time, 4 min 6 sec. For the T1-weighted spin-echo sequence, the following imaging parameters were used: 350/14; section thickness, 4 mm; intersection gap, 0.8 mm; field of view, 16 cm; matrix, 512 × 512; 2 averages; and acquisition time, 3 min 48 sec. In addition, three other sequences were performed as part of our routine hip MR arthrography protocol: a coronal T1-weighted spin-echo sequence (524/14; section thickness, 3 mm; intersection gap, 0.6 mm; field of view, 16 cm; matrix, 512 × 512; 1 average; and acquisition time, 3 min 25 sec), a coronal proton density–weighted turbo spin-echo sequence with fat saturation (2,500/42; section thickness, 3 mm; intersection gap, 0.6 mm; flip angle, 180°; field of view, 16 cm; matrix 512 × 512; turbo factor, 7; 1 average; and acquisition time, 2 min 50 sec), and an oblique axial fat-suppressed FLASH sequence (600/11.8; section thickness, 3 mm; intersection gap, 0.6 mm; flip angle, 60°; field of view, 16 cm; matrix, 512 × 512; 1 average; and acquisition time, 3 min 52 sec). Intraarticular contrast injection was performed in a standardized fashion 30 min or less before scanning: 1–2 mL of local anesthetics ([mepivacaine hydrochloride 2%] Scandicain, AstraZeneca), 1 mL of iodinated contrast agent ([iopamidol 200 mg I/mL] Iopamiro 200, Bracco), and a mean of 8 mL (range, 6–10 mL) of a diluted MR contrast agent ([gadopentetate dimeglumine] Magnevist, Schering) at a concentration of 4 mmol/L were injected under fluoroscopic control. Image Analysis All 50 MR studies were reviewed on a PACS workstation (ID.Read, Image Devices) by two musculoskeletal radiologists separately and blinded to clinical data including surgical reports. Both 3D double-echo steady-state and T1-weighted spinecho sequences were reviewed during separate sessions at least 1 week apart. During an additional session, the diagnostic performance of the coronal fat-suppressed proton density–weighted turbo spinecho sequence was evaluated in the 21 hips that underwent surgery because similar sequences are often included in MR arthrography protocols. The reviewers were asked to report whether acetabular, femoral, or both types of cartilage lesions were present.

AJR:183, December 2004

MR Arthrography of the Hip Error Analysis A board of five radiologists including the two initial reviewers reevaluated all MR studies with surgical proof (n = 21) with regard to potential sources of error. This evaluation was performed in consensus. A predefined list of sources of error was used: thin cartilage; no contrast between the two cartilage layers; anatomic irregularities at the transition between cartilage and adjacent structures (transitional zones between cartilage and acetabular labrum, acetabular fossa, fovea of the femoral head, and border of the femoral neck); partial volume effect artifacts; blurring; motion artifacts; chemical shift artifacts; noisy images; reviewer error; and small lesion (limited to a few pixels). Lesion Conspicuity Lesion conspicuity was graded between 1 and 5 by the same board of radiologists. Grades were defined as follows: 1, lesion not visible even in retrospect; 2, only a few pixels visible in retrospect; 3, lesion retrospectively visible, prospective error in diagnosis understandable; 4, lesion clearly visible but not well defined; and 5, lesion clearly visible and well defined. Statistics On the basis of the data of the 21 patients with surgical proof, sensitivity, specificity, accuracy, and positive and negative predictive values were calculated for all three evaluated sequences and for both reviewers. The mean lesion conspicuity grade (all lesions diagnosed by reviewer 1, reviewer 2, or both reviewers) was calculated for 3D double-echo steady-state and T1-weighted spinecho sequence separately. The p values were calculated using a Wilcoxon’s signed rank test. The kappa statistics were calculated for all 50 examinations to assess the level of interobserver agreement. According to Landis and Koch [32], a kappa value of 0.20 or less indicates poor agreement; 0.21–0.40, fair; 0.41–0.60, moderate; 0.61–0.80, good; and 0.81–1.0, very good agreement.

Results

At surgery, 20 acetabular and six femoral cartilage lesions were found. Cartilage surfaces of one acetabulum and 15 femoral heads were found to be intact. Results of both reviewers for all three study sequences are shown in Table 1. In general, all sequences were specific but relatively insensitive. The sensitivity of the 3D double-echo steady-state sequence was identical to that of the T1-weighted spin-echo sequence for reviewer 1, and the sensitivity for both sequences was equal for reviewer 2. The specificity of the 3D double-echo steady-state sequence was slightly lower for reviewer 1 but was higher for reviewer 2. The 3D double-echo steady-state

AJR:183, December 2004

TABLE 1

Diagnostic Performance of 3D Double-Echo Steady-State and T1-Weighted Spin-Echo Sequences in 21 Surgically Evaluated Hips (42 Cartilage Surfaces) 3D Double-Echo Steady-State

Performance

Coronal Proton Density–Weighted Fat-Saturated

T1-Weighted Spin-Echo

Reviewer 1 Reviewer 2 Reviewer 1 Reviewer 2 Reviewer 1 Reviewer 2 True-positive (no.) True-negative (no.) False-positive (no.) False-negative (no.) Sensitivity (%) Specificity (%) Accuracy (%) Positive predictive value (%) Negative predictive value (%)

15 14 2 11 58 88 69 88 56

16 15 1 10 62 94 74 94 60

sequence was less accurate than the T1weighted spin-echo sequence for reviewer 1 but was almost identical for reviewer 2. The coronal proton density–weighted turbo spinecho sequence was slightly more sensitive but was less specific and did not improve accuracy compared with the other two study sequences. The results of the error analysis are shown in Table 2. Small lesions and partial volume effect artifacts were the most common presumed reasons for false-negative diagnoses. The remaining possible sources of errors were either rare or not considered to be the main reason for false diagnoses at all. Lesion conspicuity as determined by the panel is summarized in Table 3. Significantly

TABLE 2

21 13 3 5 81 81 81 88 72

16 16 0 10 62 100 76 100 62

18 12 4 8 69 75 71 88 56

20 11 5 6 81 69 76 94 60

higher (p = 0.036, Wilcoxon’s signed rank test) average lesion conspicuity grade of the 3D double-echo steady-state sequence was found. Interobserver agreement for the detection of cartilage lesions in acetabular and femoral cartilage was fair (κ = 0.40) for the 3D double-echo steady-state images and moderate (κ = 0.55) for the T1-weighted spin-echo images.

Discussion

The double-echo steady-state sequence used in this investigation has previously been described by Hardy et al. [30] and has to date been applied on patellar [16] and femorotibial [10] cartilage. In combination with water

Error Analysis of 3D Double-Echo Steady-State Versus T1-Weighted Spin-Echo Sequences in 50 Arthrograms 3D Double-Echo Steady-State

T1-Weighted Spin-Echo

Parameters False-positive diagnosis Thin cartilage Anatomic borders of cartilage Motion artifacts Chemical shift artifacts Total False-negative diagnosis Thin cartilage Anatomic borders of cartilage Partial volume effect artifacts Motion artifacts Reviewer error Small lesions Total

Reviewer 1

Reviewer 2

Reviewer 1

Reviewer 2

1 1 — —

— — 1 —

2 — — 1

— — — —

2

1

3

0

1 1 4 — 1 4

1 1 3 — 2 3

1 — 2 1 — 1

1 — 3 1 2 3

11

10

5

10

Note.—Dash (—) indicates there was no lesion of that particular grade.

1731

Knuesel et al.

TABLE 3

Conspicuity of Cartilage Lesions on 3D Double-Echo Steady-State Versus T1-Weighted Spin-Echo Sequences in 50 MR Arthrograms 3D Double-Echo Steady-State

T1-Weighted Spin-Echo

Lesion Conspicuity True-positive diagnosis Grade 1 Grade 2 Grade 3 Grade 4 Grade 5 Total False-negative diagnosis Grade 1 Grade 2 Grade 3 Grade 4 Grade 5 Total

Reviewer 1

Reviewer 2

Reviewer 1

Reviewer 2

— 1 2 4 8

1 2 1 5 7

4 2 2 9 4

1 2 1 8 3

15

16

21

15

3 4 3 1 —

2 3 4 — 1

2 3 — — —

5 3 1 1 —

11

10

5

10

Note.—Data are numbers of cartilage areas. Average grade of conspicuity is determined by adding true-positive and falsenegative diagnoses. Dash (—) indicates there was no lesion of that particular grade. Grade 1 = lesion not visible, even in retrospect; grade 2 = only a few pixels visible in retrospect; grade 3 = lesion retrospectively visible, prospective error in diagnosis understandable; grade 4 = lesion clearly visible but not well defined; grade 5 = lesion clearly visible and well defined.

excitation, the 3D double-echo steady-state sequence has proven to be accurate in the determination of cartilage volume of elbow joints [25]. In our 21 patients with surgical confirmation, the accuracy of the 3D double-echo steady-state and that of the T1-weighted spinecho sequences were comparable for reviewer 2 (74% vs 76%, respectively). For reviewer 1, the 3D double-echo steady-state sequence was unexpectedly inferior to the T1-weighted spinecho sequence (69% vs 81%, respectively).

T1-weighted images are still included in many MR joint protocols. They often provide excellent anatomic information and show abnormalities of subchondral bone that cannot necessarily be seen on gradient-echo images. This last aspect may explain the somewhat unexpected high sensitivity of the T1-weighted sequence in detecting cartilage lesions, although section thickness, intersection gap, and field of view favor the 3D double-echo steady-state sequence. Schmid et al. [5] found that secondary signs of osteoarthri-

tis (e.g., subchondral sclerosis, subchondral cysts, and osteophytes) may be misleading. However, our data may still direct the reviewer to adjacent minor cartilage abnormalities (Figs. 2 and 3). The coronal fat-suppressed proton density–weighted turbo spin-echo sequence yields images with characteristics similar to the T1-weighted fat-suppressed spin-echo sequence, which is preferred by many radiologists for MR arthrography. However, this sequence was not superior to the other two sequences used in our study (Table 1). Contrary to the diagnostic performance that showed a slight advantage for the T1weighted sequence, lesion conspicuity for the 3D double-echo steady-state sequence was significantly superior to the T1-weighted spin-echo sequence when reviewed retrospectively (Figs. 2–4). Although this fact is not relevant with regard to the diagnostic performance of the two reviewers involved in this investigation, the 3D double-echo steady-state sequence may be valuable for less experienced radiologists or for discussion with clinicians and patients less accustomed to the relatively subtle signal differences seen on T1-weighted spin-echo images. In addition, the smaller possible slice thickness associated with the 3D acquisition used for the double-echo steady-state sequence allows reduction of partial volume effect artifacts, which may be more relevant in small bones with thin-cartilage surfaces such as in the wrist and foot than in large bones with thick-cartilage surfaces. The diagnostic performances of all the investigated sequences are comparable to the

Fig. 2.—Subchondral bone irregularity may influence detection of acetabular cartilage defect on T1-weighted spin-echo image, as shown in 21year-old man with acetabular cartilage lesion. A, Sagittal T1-weighted MR image (TR/TE, 350/14) reveals hypointensity of subchondral bone marrow (arrowheads) adjacent to small area of increased cartilage signal (arrow). B, Sagittal 3D double-echo steady-state image (24/6.5; flip angle, 25°) that corresponds to A shows cartilage defect (arrow) is more conspicuous but subchondral bone marrow abnormalities are not well seen.

A 1732

B AJR:183, December 2004

MR Arthrography of the Hip Fig. 3.—Extensive cartilage defect seen in both sequence types, as shown in 51-yearold man with cartilage lesion in acetabulum. A, Sagittal T1-weighted image (TR/TE, 350/ 14) shows well-defined cartilage loss in anterior acetabulum (between arrows). Obvious adjacent bone marrow abnormality (arrowheads) is visible. B, Sagittal 3D double-echo steady-state image (24/6.5; flip angle, 25°) that corresponds to A shows size of cartilage lesion (between arrows) is identical to that shown in A. Subchondral bone marrow appears nearly normal.

A

B

A

B

Fig. 4.—Extensive cartilage defects seen on both 3D double-echo steady-state and T1-weighted spin-echo images, as shown in 33-year-old woman with acetabular and femoral articular cartilage damage. A, Sagittal T1-weighted image (TR/TE, 350/ 14) shows signal irregularities within acetabular (curved black arrow), posterior femoral (straight white arrows), and anterior femoral (straight black arrow) cartilage. Subchondral bone cysts (curved white arrows) and sclerosis (arrowheads) are visible on acetabular side. B, Sagittal 3D double-echo steady-state image (24/6.5; flip angle, 25°) that corresponds to A renders posterior acetabular (curved black arrows), anterior femoral (straight black arrows), and posterior femoral (straight white arrows) cartilage defects more conspicuous than A. In this patient, subchondral cysts (curved white arrows) and adjacent cartilage defects are better seen on 3D double-echo steady-state image than on T1-weighted image.

results of previously published studies of the hip and other joints. In an MR arthrography study of the hip, accuracies of 78% (reviewer 1) and 69 % (reviewer 2) for detecting cartilage defects were found. A 1.0-T scanner and three different imaging planes (a sagittal T1weighted turbo spin-echo, an oblique coronal T1-weighted turbo spin-echo and oblique coronal T1-weighted fat-suppressed spinecho, and an oblique axial T1-weighted fatsuppressed FLASH sequence) were used [5]. Nakanishi et al. [15] investigated articular cartilage of the hip without intraarticular contrast administration. They used a dedicated fat-suppressed 3D SPGR sequence during longitudinal leg traction. In 14 of

AJR:183, December 2004

their 15 patients with osteonecrosis (n = 5), acetabular dysplasia (n = 5), or osteoarthritis (n = 5), cartilage abnormalities corresponded well to macroscopic or arthroscopic findings. The MR appearance of articular cartilage was normal in all eight young volunteers included in that study [15]. Hodler et al. [28] reported a reasonable correlation (r = 0.25–0.58) between anatomic and MR measurements of cartilage thickness in both the acetabulum and femur. A fat-suppressed SPGR sequence and a T1-weighted spin-echo sequence were used in that investigation. Several studies performed in the knee with fat-suppressed 3D SPGR [7, 12–15, 23], FLASH [8, 18], T2- and intermediate-

weighted fat-suppressed fast spin-echo [6, 9], and 3D double-echo steady-state [10, 16] sequences have reported sensitivities, specificities, and accuracies of 43–94%, 88–100%, and 83–98%, respectively. These values cannot directly be compared with values obtained in the hip, however. Hip cartilage is as thin as 1–2 mm (mean) [28, 33, 34] versus knee cartilage that is reported to be up to 7 mm thick at the patella [33]. In addition, hip cartilage is less accessible for imaging because of its anatomic location deep within soft tissue. Flexible surface coils or body array coils are commonly used for imaging of the hip instead of dedicated send–receive extremity coils with their more homogeneous

1733

Knuesel et al. field that are commonly used for imaging the knee. Previously published MR studies of joints with relatively thin cartilage have shown sensitivities, specificities, and accuracies similar to our results in the glenohumeral joint (53–100%, 51–87%, 65–75%, respectively) [20] and ankle joint (59–80%, 76–97%, 75–88%, respectively) [24]. Interobserver agreement was fair to moderate in our study (κ = 0.40 for 3D double-echo steady-state and κ = 0.55 for T1-weighted spin-echo sequences). Similar or inferior kappa values have previously been reported for patellar cartilage (κ = 0.45 for 3D doubleecho steady-state images and κ = 0.46 for T1weighted turbo spin-echo images [16]), for the hip (κ = 0.20–0.31 for MR arthrography and standard spin-echo and 2D fat-suppressed FLASH sequences [5]), and for the glenohumeral joint (κ = 0.20–0.27 for T1-weighted spin-echo, intermediate-, and T2-weighted fast spin-echo sequences [20]). The performance of standard MRI and MR arthrography also must be compared with CT arthrography, which may be used as an alternative imaging technique for cartilage imaging. CT arthrography has successfully been used in the knee [14, 17, 18, 27] and ankle [24] joints. Vande Berg et al. [17] compared the performance of CT arthrography and nonarthrographic MRI in the detection of cartilage lesions in cadavers. They found sensitivities and specificities of 80– 88% and 78–85%, respectively. In a study of cadaveric specimens, Rand et al. [14] found that the sensitivity of CT arthrography and that of standard MR sequences (T1- and proton density–weighted) were identical in the detection of different degrees of abnormalities of the cartilage surface (86% for all methods) and were slightly lower than in MR arthrography (90% for SPGR sequence). However, CT arthrography failed to depict intrachondral lesions. Gagliardi et al. [27] reported sensitivities, specificities, and accuracies for MR arthrography (80–100%, 98–100%, 94– 100%, respectively) that are superior to those of CT arthrography (73–100%, 100%, 92–100%, respectively) in the diagnosis of chondromalacia patellae (grade 2 or higher) in vivo . In another study comparing CT arthrography and MR arthrography of the ankle, similar sensitivities, specificities, and accuracies were found for the two methods [24] (tibia and talus: 60–81%, 89–97%, 90–94% vs 59– 80%, 76–97%, 75–88%, respectively). In the hip, CT arthrography is associated with a relevant amount of radiation exposure. This is

1734

important because cartilage damage in the hip joint occurs in young individuals. Therefore, MR examinations should be favored in this region. The comparison of two sequences in only one imaging plane as used in our investigation is somewhat artificial because consultation of coronal images may prevent false results that are due to partial volume effect artifacts or problems relating to anatomic transitional zones (e.g., the acetabular cartilage abutting the labrum). However, in a study design with additional imaging planes, any differences between the two evaluated types of sequences might have become less evident. In conclusion, for MR arthrograms, both a standard T1-weighted spin-echo and a waterexcitation 3D double-echo steady-state MR sequence are specific but insensitive to cartilage lesions of the hip joint. The 3D double-echo steady-state sequence optimized for cartilage imaging does not improve the diagnostic performance in this setting, although conspicuity of cartilage lesions was markedly better than on T1-weighted spin-echo sequence. References 1. Ito K, Minka MA 2nd, Leunig M, Werlen S, Ganz R. Femoroacetabular impingement and the cameffect: a MRI-based quantitative anatomical study of the femoral head-neck offset. J Bone Joint Surg Br 2001;83:171–176 2. Ganz R, Gill TJ, Gautier E, Ganz K, Krügel N, Berlemann U. Surgical dislocation of the adult hip. J Bone Joint Surg Br 2001;83:1119–1124 3. Jackson DW, Scheer MJ, Simon TM. Cartilage substitutes: overview of basic science and treatment options. J Am Acad Orthop Surg 2001;9:37– 52 [Erratum in J Am Acad Orthop Surg 2001;9:147] 4. Nehrer S, Minas T. Treatment of articular cartilage defects. Invest Radiol 2000;35:639–646 5. Schmid MR, Nötzli HP, Zanetti M, Wyss TF, Hodler J. Cartilage lesions in the hip: diagnostic effectiveness of MR arthrography. Radiology 2003;226:382–386 6. Bredella MA, Tirman PFJ, Peterfy CG, et al. Accuracy of T2-weighted fast spin-echo MR imaging with fat saturation in detecting cartilage defects in the knee: comparison with arthroscopy in 130 patients. AJR 1999;172:1073–1080 7. Disler DG, McCauley TR, Kelman CG, et al. Fatsuppressed three-dimensional spoiled gradientecho MR imaging of hyaline cartilage defects in the knee: comparison with standard MR imaging and arthroscopy. AJR 1996;167:127–132 8. Recht MP, Piraino DW, Paletta GA, Schils JP, Belhobeck GH. Accuracy of fat-suppressed threedimensional spoiled gradient-echo FLASH MR imaging in the detection of patellofemoral articular cartilage abnormalities. Radiology 1996;198: 209–212

9. Broderick LS, Turner DA, Renfrew DL, Schnitzer TJ, Huff JP, Harris C. Severity of articular cartilage abnormality in patients with osteoarthritis: evaluation with fast spin-echo MR vs arthroscopy. AJR 1994;162:99–103 10. Murphy BJ. Evaluation of grades 3 and 4 chondromalacia of the knee using T2*-weighted 3D gradient-echo articular cartilage imaging. Skeletal Radiol 2001;30:305–311 11. Hodler J, Trudell D, Pathria MN, Resnick D. Width of the articular cartilage of the hip: quantification by using fat-suppression spin-echo MR imaging in cadavers. AJR 1992;159:351–355 12. Brossmann J, Preidler KW, Daenen B, et al. Imaging of osseous and cartilaginous intraarticular bodies in the knee: comparison of MR imaging and MR arthrography with CT and CT arthrography in cadavers. Radiology 1996;200:509–517 13. Recht MP, Kramer J, Marcelis S, et al. Abnormalities of articular cartilage in the knee analysis of available MR techniques. Radiology 1993;187: 473–478 14. Rand T, Brossmann J, Pedowitz T, Ahn JM, Haghigi P, Resnick D. Comparison of conventional MR imaging and MR and CT arthrography in cadavers. Acta Radiol 2000;41:492–497 15. Nakanishi K, Tanaka H, Nishii T, Masuhara K, Narumi Y, Nakamura H. MR evaluation of the articular cartilage of the femoral head during traction: correlation with the resected femoral head. Acta Radiol 1999;40:60–63 16. Ruehm S, Zanetti M, Romero J, Hodler J. MRI of patellar articular cartilage: evaluation of an optimized gradient echo sequence (3D DESS). J Magn Reson Imaging 1998;8:1246–1251 17. Vande Berg BC, Lecouvet FE, Poilvache P, et al. Assessment of knee cartilage in cadavers with dual-detector spiral CT arthrography and MR imaging. Radiology 2002;222:430–436 18. Daenen BR, Ferrara MA, Marcelis S, Dondelinger RF. Evaluation of patellar cartilage surface lesions: comparison of CT arthrography and fat-suppressed FLASH 3D MR imaging. Eur Radiol 1998;8:981–985 19. Horii M, Kubo T, Hirasawa Y. Radial MRI of the hip with moderate osteoarthritis. J Bone Joint Surg Br 2000;82:364–368 20. Guntern DV, Pfirrmann CWA, Schmid M, et al. Articular cartilage lesions of the glenohumeral joint: diagnostic effectiveness of MR arthrography and prevalence in patients with subacromial impingement syndrome. Radiology 2003;226: 165–170 21. Peterfy CG, van Dijke CF, Lu Y, et al. Quantification of the volume of articular cartilage in the metacarpophalangeal joints of the hand: accuracy and precision of three-dimensional MR imaging. AJR 1995;165:371–375 22. Disler DG, Recht MP, McCauley TR. MR imaging of articular cartilage. Skeletal Radiol 2000;29: 367–377 23. Disler DG, McCauley TR, Wirth CR, Fuchs MD. Detection of knee hyaline cartilage defects using fat-suppressed three-dimensional spoiled gradient-echo MR imaging: comparison with standard MR imaging and correlation with arthroscopy. AJR 1995;165:377–382

AJR:183, December 2004

MR Arthrography of the Hip 24. Schmid MR, Pfirrmann CWA, Hodler J, Vienne P, Zanetti M. Cartilage lesions in the ankle joint: comparison of MR arthrography and CT arthrography. Skeletal Radiol 2003;32:559–565 25. Graichen H, Springer V, Flaman T, et al. Validation of high-resolution water-excitation magnetic resonance imaging for quantitative assessment of thin cartilage layers. Osteoarthritis Cartilage 2000;8:106–114 26. Bachmann G, Heinrichs C, Jurgensen I, Rominger M, Scheiter A, Rau WS. Comparison of different MR techniques in the diagnosis of degenerative cartilage diseases: in vitro study of 50 joint specimens of the knee at T1.5 [in German]. Rofo Fortschr Geb Rongenstr Neuen Bildgeb Verfahr 1997;166:429–436

27. Gagliardi JA, Chung EM, Chandnani VP, et al. Detection and staging of chondromalacia patellae: relative efficacies of conventional MR imaging, MR arthrography, and CT arthrography. AJR 1994;163:629–636 28. Hodler J, Trudell D, Pathria MN, Resnick D. Width of the articular cartilage of the hip: quantification by using fat-suppression spin-echo MR imaging in cadavers. AJR 1992;159:351–355 29. McGibbon CA, Dupuy DE, Palmer WE, Krebs DE. Cartilage and subchondral bone thickness distribution with MR imaging. Acad Radiol 1998;5:20–25 30. Hardy PA, Recht MP, Piraino D, Thomasson D. Optimization of a dual echo in the steady state (DESS) free-precession sequence for imaging

31.

32.

33.

34.

cartilage. J Magn Reson Imaging 1996;6:329– 335 Mosher TJ, Pruett SW. Magnetic resonance imaging of superficial cartilage lesions: role of contrast in lesion detection. J Magn Reson Imaging 1999;10:178–182 Landis JR, Koch GC. The measurement of reviewer agreement for categorical data. Biometrics 1977;33:159–174 Adam C, Eckstein F, Milz S, Putz R. The distribution of cartilage thickness within the joints of the lower limb of elderly individuals. J Anat 1998;193:203–214 Shepherd DET, Seedhom BB. Thickness of human articular cartilage in joints of the lower limb. Ann Rheum Dis 1999;58:27–34

The ARRS 2005 annual meeting categorical course will be on cardiopulmonary imaging.

AJR:183, December 2004

1735