| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Musculoskeletal Imaging |
1 From the Departments of Diagnostic Radiology (K.H.C., B.H.P.) and Orthopedic Surgery (D.C.L., S.D.K.), Yeungnam University Medical Center, College of Medicine, 317-1 Daemyung-Dong, Nam-Ku, Taegu 705-717, Korea; the Department of Diagnostic Radiology, National University Hospital, Faculty of Medicine, National University of Singapore (R.K.C.); the Department of Diagnostic Radiology, Henry Ford Hospital, Detroit, Mich (J.A.B.); and the Department of Diagnostic Radiology, St-Luc Hospital, Montreal, Quebec, Canada (E.C.). Received August 6, 1998; revision requested September 11; final revision received September 26, 2000; accepted October 11. K.H.C. supported in part by the Chunma Foundation for Medical Science. Address correspondence to K.H.C. (e-mail: khcho@medical.yeungnam.ac.kr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: US images were obtained in 30 knees in 15 asymptomatic volunteers as a control group and in 35 patients clinically suspected of having an acute PCL injury. Only the distal half of the PCL was evaluated. Of the 35 patients, 28 had their PCL status confirmed: 13 had a normal PCL at magnetic resonance (MR) imaging plus clinical examination, and 15 had a torn PCL at either MR imaging and surgery or MR imaging and clinical follow-up.
RESULTS: Normal PCLs were homogeneously hypoechoic, with a well-defined posterior border. Torn PCLs were heterogeneously hypoechoic (12 [80%] of 15 patients), with an indistinct posterior margin (11 [73%] of 15 patients). Torn PCLs were significantly thicker (range, 12.0–20.0 mm; mean, 15.6 mm ± 2.5 [SD]; P < .01), as compared with normal PCLs in 13 patients (range, 3.8–5.8 mm; mean, 4.6 mm ± 1.0; P < .01) and in the volunteers (range, 3.7–6.2 mm; mean, 4.5 mm ± 1.2; P < .01).
CONCLUSION: An acutely torn PCL thickens (>10 mm), loses its sharply defined posterior border, and has a heterogeneously hypoechoic appearance. US may be useful as a screening examination for patients suspected of having PCL injury and for deciding whether to perform more expensive MR imaging or surgical intervention.
Index terms: Knee, injuries, 4526.1298, 4526.4857 • Knee, ligaments, menisci, and cartilage, 4526.1298, 4526.4857 • Knee, US, 4526.1298, 4526.4857
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Arthroscopy, known as the reference standard for diagnosis, is invasive and requires multiple portals for three-zone concept evaluation of the entire length of the PCL (2–4,6,7). Magnetic resonance (MR) imaging is known to be as useful as arthroscopy but is expensive (4,8,9). In the clinical setting there is a need for an inexpensive, noninvasive, sensitive, and specific screening examination to evaluate the PCL.
Ultrasonography (US) has not been universally performed to evaluate the PCL, even though it does not cost as much as MR imaging and is performed more rapidly. Also, US requires neither the anesthesia nor the aseptic procedures that are prerequisites for arthroscopy. Concomitant injuries of the knee are common with PCL injuries, and this would make US a less ideal imaging choice. Little has been written with regard to US evaluation of the PCL (10–14). We performed US in normal or acutely torn PCLs to describe the findings and evaluate the usefulness of US in the injured PCL.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
We elected to use a linear-array transducer because it provides precise contour definition and maximum echogenicity of the PCL, which cannot be obtained with a curvilinear transducer. The PCL was investigated by means of a posterior approach by applying the transducer over the popliteal fossa. US was performed by using a 5–10-MHz broadband linear-array transducer (HDI or HDI-3000; ATL, Bothell, Wash), with the volunteers and patients lying prone with the knee in the most comfortable neutral position. The focus of the transducer was adjusted at the level of the ligament, without using a stand-off pad. The transducer was longitudinally positioned at the site of the distal attachment of the PCL, in the posterior intercondylar area of the proximal tibia (Fig 1). The proximal corner of the transducer was pushed down as much as possible in a "heel-toe" maneuver and was tilted slightly to achieve a plane parallel to the long axis of the ligament. The proximal half of the PCL was too deep to be depicted. The ligament’s distal half, which ran parallel down the posterior intercondylar area, was well demonstrated as a configuration that thickened proximally and tapered distally. With US, we could not discriminate the PCL from the adjacent meniscofemoral ligaments. Transverse images were not obtained because it was difficult to standardize the level at which the images were obtained.
|
|
We could not find any previously published references for the PCL thickness (anteroposterior dimension) at the level of the tibial spine or for methods of US measurement of PCL thickness. We thus decided to measure the PCL thickness at the level of the tibial spine five times in each knee and average the middle three measurements after eliminating the longest and shortest. The results of this examination provided information about the normal morphology and echogenicity of the PCL, as well as about the thickness of the normal ligament, to serve as a basis for comparison with US findings in injured PCLs. MR imaging was not performed in the asymptomatic volunteers.
Patient Group
From October 1995 to December 1996, US examinations of the PCL were performed by a single author (K.H.C.) in the 35 patients who were clinically suspected of having an acute PCL tear. The study was approved by Yeungnam University Medical Center’s institutional review board. Written informed consent was obtained from the patients or their guardians. We evaluated the same items as we did in the control group with regard to US findings, and the same type of equipment was used.
Of the 35 patients, 28 (18 men and 10 women; age range, 18–65 years; mean age, 42 years) received confirmation of the status of their PCL. Thirteen of these 28 patients were confirmed as having a normal PCL in accordance with MR imaging findings plus clinical data and did not undergo surgical intervention. Another 13 patients were proved to have a torn PCL at MR imaging and orthopedic surgery (performed by D.C.L. and S.D.K.) within 3 days after the imaging studies. The remaining two patients were confirmed as having a torn PCL at MR imaging and clinical follow-up.
No patient had bilateral PCL injuries. Ten patients who had difficulty lying prone were examined in the lateral decubitus position, with the injured knee positioned nondependently. Examination of the contralateral uninjured knee in the same patient was attempted routinely for comparison because of our limited experience. However, examination of the contralateral knee was precluded in the 10 patients, who were unable to stay in the prone position or change the leg position because of pain. Also, it was difficult to examine the asymptomatic PCL of the contralateral knee in the dependent position with clinical expedience. We examined only the PCL whether the patients had combined injuries or not.
In the patient group, MR imaging was performed with one of three units (Gyroscan [0.5 T], Philips Medical Systems, Eindhoven, the Netherlands; Signa [1.5 T], GE Medical Systems, Milwaukee, Wis; or Magnetom Vision [1.5 T], Siemens, Erlangen, Germany) by using the known standard protocol for each imager. MR imaging was performed within 24 hours before or after US; most examinations were performed on the same day. The MR imaging protocol included T1- (repetition time msec/echo time msec, 475–650/12–25) and T2-weighted (2,480–5,000/19–25, 80, or 102–119) imaging in coronal and sagittal planes, and T1-weighted imaging (475–650/12–25) in the transverse plane. Studies were performed with a 16–20-cm field of view; a 4–5-mm section thickness; a 0.3–1.0-mm intersection gap; and a 179 x 256, 192 x 256, or 270 x 512 matrix, depending on the MR imager used for each patient.
We diagnosed the PCL as normal on MR images when it appeared smooth, with a regularly thick configuration and a well-defined band of low signal intensity on T1- and T2-weighted images, without abnormal high signal intensity. The MR imaging criteria for a torn PCL consisted of failure to identify the PCL, amorphous areas of high signal intensity in the region of the PCL on T1- and T2-weighted images, loss of continuity of the ligament, or focal discrete disruption of all visible fibers (8). The MR images were interpreted independently at first by two musculoskeletal radiologists (R.K.C., J.A.B.) who were blinded to the US findings in each case. They disagreed on two normal and two abnormal cases. These disagreements were resolved by means of consensus after discussion with a third musculoskeletal radiologist (K.H.C.), although the third radiologist was not blinded to either MR imaging or US findings. However, the disagreements on the two normal cases were resolved with clinical follow-up, and the two abnormal cases were confirmed surgically as tears.
Statistical Evaluation
Statistical analysis was performed (SPSS version 7.5 for Windows; SPSS Korea, Seoul). By performing the Student t test, we compared the thickness of normal PCLs between the control and patient groups at US with that of the torn ligaments in the patient group.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
In the 15 patients with a torn PCL, surgery and/or MR imaging showed that three patients had tears in the proximal third; seven, in the middle third; and five, in the distal third. At US, the thickness of the torn PCL was increased regardless of where the tear occurred.
Hematoma formation adjacent to the PCL was seen at US in three of the 15 patients with PCL tears (Fig 3). A transverse hypoechoic defect representing the tear was demonstrated at US in four of the 15 patients (Fig 4).
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The most commonly performed clinical examination for assessing the status of the PCL is the posterior drawer test (2,3,7,9). The clinical sign is not positive in all patients who have an acute PCL disruption because of the presence of muscle spasm and a tense hemarthrosis (2). Hughston et al (5) reported that posterior drawer test results before anesthesia were negative in 26 of 29 knees with acute PCL injuries and were still negative or equivocal in 20 knees after anesthesia. Patten et al (6) found that initial clinical assessment in 10 of 32 patients who had acute PCL tears could not be performed adequately. In another six patients, they noted that PCL injury was not diagnosed initially, despite adequate clinical examination (6). The positive drawer sign is variably evident in 50%–90% of PCL injuries (4,9,16). False-negative posterior drawer test results may be related to resistance from an intact arcuate complex and from the combined posterolateral corner ligament (3).
Injuries commonly associated with PCL disruption involve the anterior cruciate ligament (65%), medial collateral ligament (50%), and medial meniscus (30%) (16). The most common site of PCL tear varies; in a series of 59 acute PCL tears, the PCL injury sites were tibial (42%), femoral (36%), and midsubstance (22%) (16). In another study (8), 45 (63%) of 71 patients who had MR imaging findings of PCL tear had midsubstance injuries, 19 (27%) had proximal injuries, and seven (10%) had distal ligament disruptions. The authors of a third report (6) pointed out that abnormal signal intensity on MR images was localized to the midportion of the PCL in 66%, to the proximal portion in 9%, to the distal portion in 19%, and to the entire length of the ligament in 6%. In the current study, MR imaging and/or surgery confirmed that most cases were tears in the distal two-thirds of the PCL; these probably were due to injuries sustained in car accidents (8).
The PCL anatomically originates from the lateral surface of the medial femoral condyle and inserts onto the intercondylar area of the proximal tibia 1 cm below the articular surface. It runs in an oblique-sagittal course with the proximal portion of the PCL, which is deviated medially and angled forward 30°–45° in the sagittal plane (4,8,17–19). The distal half of the PCL is in a plane parallel and close to the posterior half of the proximal tibia (4,8).
The PCL has a mean length of 38 mm and a mean width (mediolateral dimension) of 13 mm (19). The cross-sectional area of the PCL decreases from the proximal to the distal portion (20). Our experience was that at US, the thickness (anteroposterior dimension) of the normal PCL at the level of the tibial spine was 3.7–6.2 mm in the volunteer group and 3.8–5.8 mm in the contralateral asymptomatic knee in the patient group. The thickness of the PCL in the healthy volunteers in our study was less than that reported in previous studies (6,8) in which the normal PCL was described as being 10–15 mm thick, although the authors did not specify the level of measurement and whether the dimension was anteroposterior or mediolateral. This difference in thickness between our results and those in other studies (6,8) may be explained by the fact that the PCL is narrowest in its middle portion and fans out proximally and distally (19,20). The proximally fanned-out portion of the PCL attaches to the femur with a mean length of 32 mm in an anteroposterior direction, whereas on the tibia, the PCL attaches with a mean width of 13 mm in a mediolateral direction (19).
On the other hand, our US results for the thickness of the PCL may not reflect the true thickness because the US measurement method we used may include fat and other soft tissue between the proximal tibia and the PCL, although the PCL thickness in our study was thinner than that in other studies (6,8). Meniscofemoral ligaments may also have inadvertently been included in the measurements in some cases, because these may cross the PCL anteriorly and posteriorly at the area measured.
Suzuki et al (10) reported that the normal PCL was highly echoic at US, the torn PCL lost its tension and hyperechoic appearance, and the entire length of the PCL was demonstrated in a posterior US approach (5.0 or 7.5 MHz). In our study, however, only the distal half of the PCL could be depicted with a posterior approach because the proximal segment of the PCL is too deep to be scanned and changes its course in a direction transverse to the sound beam, which makes it invisible.
In our experience, the PCL was hypoechoic in comparison with that in the study by Suzuki et al (10) because of an anisotropic artifact from the obliquity of its course to an insonated ultrasound beam. On the other hand, the PCL in Suzuki et al’s (10) figures seems, in our opinion, to be posterior intraarticular fat rather than the PCL itself. In that article (10), the hypoechoic triangular structure between the proximal tibia and the intraarticular fat appears to be the PCL, because the PCL in its distal half anatomically courses close to the proximal tibia.
Thickening of the injured PCL with a poorly defined posterior border, as compared with the normal ligament, is probably due to edema, hemorrhage, and fluid collection in and around the ligament (21). In our study, the thickness of the torn PCLs increased regardless of whether the PCL tear occurred proximally or distally. It could have been related to intrasubstance edema and extension of fluid and hemorrhage within the PCL sheath.
There were a number of limitations in the current study: (a) There was a patient selection bias because all included patients had a clinical likelihood of PCL injury. (b) No control patient group was included. We did not analyze other synchronous injuries of the knee, such as tear of the anterior cruciate ligament, which may result in a false-positive US diagnosis of PCL injury. (c) There was no inter- or intraobserver reliability and no assessment of accuracy, because only one nonblinded author performed US for suspected abnormal cases. (d) US was limited in direct depiction of the tear at the proximal half of the ligament. (e) We did not perform transverse US in the volunteers. (f) We were unable to differentiate a partial tear from a full-thickness tear with US. (g) Technically normal PCLs appeared hypoechoic because of an anisotropic artifact. In addition, from a technical point of view in our experience, the popliteal artery may interfere with optimal US imaging in some cases in which transducer position cannot be stabilized because of bounding arterial pulsation.
The torn PCL increases in thickness, loses its sharply defined posterior border, and has a heterogeneous hypoechoic appearance. The increase in the thickness of the ligament is the most important criterion among the US findings in the torn PCL. On the basis of our results, the thickness of the PCL at a cutoff value of 10 mm had a sensitivity and specificity of 100%, although our subjects were few and our study had some limitations.
We conclude that US may be a useful examination in patients suspected of having a PCL injury and may be helpful in making the decision to perform more costly MR imaging or surgical intervention. However, our work is preliminary, so further comparative studies with MR imaging, arthrotomy, or arthroscopy are needed.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, K.H.C.; study concepts, K.H.C.; study design, K.H.C., D.C.L., S.D.K., B.H.P.; literature research, K.H.C., D.C.L., R.K.C., J.A.B.; clinical studies, K.H.C., D.C.L., S.D.K.; data acquisition, K.H.C., D.C.L., J.A.B., S.D.K.; data analysis, K.H.C., D.C.L., R.K.C., J.A.B.; statistical analysis, K.H.C., E.C.; manuscript preparation, K.H.C., D.C.L.; definition of intellectual content, all authors; manuscript editing, K.H.C.; manuscript review, S.D.K., R.K.C., J.A.B., E.C., B.H.P.; manuscript final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. Rodriguez Jr., E. N. Vinson, C. A. Helms, and A. P. Toth MRI Appearance of Posterior Cruciate Ligament Tears Am. J. Roentgenol., October 1, 2008; 191(4): W155 - W159. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) that represents hemorrhage within the intracapsular fat (F) is contiguous with the PCL (L) through the interrupted posterior border (arrowheads). The ligament is heterogeneously hypoechoic. The hemorrhage could be expressed in and out of the tear at graded compression with the transducer. (b) Sagittal T2-weighted MR image (2,480/80) in the same patient as in a shows a normal anterior cruciate ligament (A). The PCL (L) is abruptly interrupted at the distal third (arrows), and the hemorrhage and inflammation extend into the posterior intraarticular fat (F).
