Our aim was to determine the clinical value of MRI and CT arthrography in predicting the presence of loose bodies in the elbow.
A series of 26 patients with mechanical symptoms in the elbow had plain radiography, MRI and CT arthrography, followed by routine arthroscopy of the elbow. The location and number of loose bodies determined by MRI and CT arthrography were recorded. Preoperative plain radiography, MRI and CT arthrography were compared with arthroscopy.
Both MRI and CT arthrography had excellent sensitivity (92% to 100%) but low to moderate specificity (15% to 77%) in identifying posteriorly-based loose bodies. Neither MRI nor CT arthrography was consistently sensitive (46% to 91%) or specific (13% to 73%) in predicting the presence or absence of loose bodies anteriorly. The overall sensitivity for the detection of loose bodies in either compartment was 88% to 100% and the specificity 20% to 70%. Pre-operative radiography had a similar sensitivity and specificity of 84% and 71%, respectively.
Our results suggest that neither CT arthrography nor MRI is reliable or accurate enough to be any more effective than plain radiography alone in patients presenting with mechanical symptoms in the elbow.
Arthroscopic surgery is highly effective for the removal of loose bodies in patients presenting with mechanical symptoms in the elbow.1,2 Patients who benefit the least from this procedure have pain but negative radiological findings and clinical examination.1 To avoid unnecessary surgery, it would be advantageous to use an imaging technique which is able to diagnose accurately and locate loose bodies. Plain radiography, MRI and CT arthrography are often used but there is limited information about the accuracy of these investigations.3-7 Our aim, therefore, was to determine the usefulness of these techniques in the evaluation of loose bodies in the elbow.
Patients and Methods
We studied 26 patients (25 men and one woman) with a mean age of 36 years (15 to 59). All had a history of mechanical symptoms such as locking or catching, prompting surgical treatment. They all had pre-operative plain radiography (anteroposterior and lateral), MRI and CT arthrography. The study was approved by the Human Ethics Board of the University of Western Ontario.
All MRI studies were performed on a 1.5 Tesla magnet (Siemens, Erlanger, Germany) using a flexible 15 cm surface coil. The patients' arm was placed at their side with the hand in a neutral position stabilised by a foam support. The following images were produced using 3D gradient echo (FISP; TR 30, TE 10, flip angle 55°) T2-weighted with a slice thickness of 1.5 mm reconstructed in all three planes, axial and sagittal T1 (TR 720, TE 15 slice thickness 3.0 mm) and axial proton and weighted T2 (TR 2400, TE 20/80) using a slice thickness of 3.0 mm.
All CT arthrography studies were performed on a helical scanner (General Electric Medical, Milwaukee, Wisconsin). The patient's arm was placed above their head with the patient in a prone position on the CT table. Images were performed helically in an axial plane with a slice thickness of 1.0 mm and a pitch ratio of 1:1. The CT studies were performed within 30 minutes of a double-contrast arthrogram for which air and a diluted iodinated contrast were injected into the elbow through a lateral approach. The three sets of images were reported by three radiologists (WR, GG, JB) who were aware of the clinical diagnosis but did not know the result of the arthroscopy. They recorded the location and the number of loose bodies identified.
Routine arthroscopy of the elbow was performed by two authors (GJWK, SDP) at a mean interval of 2.4 ± 1.8 months (1 to 8) after imaging. Using multiple portals, care was taken to evaluate fully all the compartments of the elbow.
The results of the MRI and the CT arthrography were compared with each other. The agreement between the three radiologists with respect to the number of loose bodies which they observed on both MRI and CT arthrography was determined using an intraclass correlation coefficient (ICC).8 The results of the MRI and the CT arthrography were then compared with the arthroscopic findings. The agreement on the number of loose bodies reported by the three radiologists on MRI and CT arthrography was compared with that determined by arthroscopy and expressed using an ICC.8 The sensitivity and specificity of plain radiography, CT arthrography and MRI for the overall presence or absence of a loose body in either compartment were compared with those of arthroscopy. Kappa values were also calculated for accuracy to reflect chance corrected agreement.9
There was little difference between CT arthrography and MRI in regard to agreement between radiologists (Table I). There was more agreement when assessing the posterior (ICC = 0.72) than the anterior compartment (ICC = 0.52 for CT arthrography and 0.41 for MRI). This would suggest that neither MRI nor CT arthrography is better than the other in terms of reliability or accuracy.
The agreement on the number of loose bodies observed on MRI and CT arthrography compared with arthroscopy was also greater in the posterior compartment (Table II). If 0.75 was taken as a benchmark of excellent reliability,8 observer 1 was able to attain this level for both CT arthrography and MRI in the posterior compartment whereas observer 2 was unable to do so with either test and observer 3 did so for MRI only. None of the anterior compartment ICCs reached this level for any radiologist.
The superior diagnostic accuracy for the posterior compartment was also demonstrated when the data were expressed as the presence or absence of loose bodies rather than the number present. The Kappa values were statistically better than chance agreement for the posterior compartment but not the anterior compartment (Table III). When these data were simplified to determine the presence or absence of loose bodies anywhere in the joint, Kappa values tended to be midway between those for the anterior and posterior compartment since the error in the anterior compartment adversely affected overall reliability.
When the diagnostic accuracy was presented in terms of sensitivity and specificity, it was evident that most of the errors in the posterior compartment were false-positive rather than false-negative. In the anterior compartment neither sensitivity nor specificity was consistently high. When data were collapsed to the presence or absence of loose bodies anywhere in the elbow, sensitivity was very high but specificity tended to be low, suggesting that false-positive results were a clinical concern (Table IV). There was no clinically relevant difference between CT arthrography and MRI in terms of the detection of loose bodies and they were no better than the original pre-operative radiological reports.
Pain in the elbow and mechanical symptoms such as 'locking' and 'catching' are often caused by loose bodies which typically are the result of osteoarthritis, osteochondritis dessicans, synovial osteochondromatosis or osteochondral fractures. To avoid unnecessary surgery, an ideal pre-operative imaging study would predict, accurately and consistently, the presence or absence of loose bodies in both the anterior and posterior compartments.
Plain radiography is readily available and inexpensive but is often inaccurate.1,3,4 Small, or primary chondral, lesions are not seen and it is difficult to distinguish between osteophytes and loose bodies. Additionally, it is difficult to determine whether the loose body is within the joint. In 1986, Boe3 reported false-positive radiographs in 12 of 35 patients and false-negative radiographs in two. Ward et al4 reported a sensitivity of 79%, a specificity of 69% and a diagnostic accuracy of 75% for plain radiography of loose bodies in the elbow. In our study, a comparison of plain radiography with arthroscopy yielded a sensitivity of 84% and a specificity of 71%. Plain radiography had a similar sensitivity and specificity to CT arthrography and MRI. Our results suggest that plain radiography alone is sufficient when assessing a patient presenting with mechanical symptoms and pain in the elbow.
CT arthrography has been commonly used and can show osteochondral fragments as small as 3 mm in diameter. It has several disadvantages including increased expense, risk of infection, potential allergic reaction to contrast dye, misinterpretation because of air bubbles and the requirement of a skilled radiologist to perform and interpret the investigation. A review of the literature yielded no studies reporting the accuracy of CT arthrography in diagnosing loose bodies. Our study found that CT arthrography was not accurate or reliable enough to predict the location or the number of loose bodies.
MRI has become popular for diagnosing loose bodies in the elbow.6,7 The benefits include multiple plane imaging, the avoidance of radiation and the diagnosis of primarily cartilaginous lesions and other lesions such as osteochondritis dessicans. MRI is expensive and observers have difficulty in differentiating between osteophytes and loose bodies.7 Quinn et al7 reported a sensitivity of 100% and a specificity of 67%. We report excellent sensitivity but a low to moderate specificity for the posterior compartment. This would suggest an increased risk of false-positive results. MRI of the anterior compartment was not consistently sensitive or specific. Our findings show that MRI in the absence of contrast is not accurate or reliable enough to predict the presence or absence of loose bodies in the elbow.
A number of limitations should be considered when interpreting the results of our study. Problems with arthroscopy as the technique of choice include the possibility of missed loose bodies and the difficulty in differentiating between osteophytes and loose bodies.
Loose bodies may have moved from one compartment to another during all of the investigations. These limitations would have applied equally across imaging modalities and therefore were unlikely to have contributed any bias to our conclusions. Another limitation was the relatively small sample size which resulted in statistical comparisons which were underpowered. Studies with larger sample sizes may establish statistical differences between CT arthrography, MRI and plain radiography, but also need to determine whether observed differences fall within a clinically meaningful range. Finally, since we carried out the study, advances have been made in both CT and MRI. Currently at our centre, CT arthrography is performed with a singlecontrast technique, thinner slices and multiplanar reconstruction. MRI is done with thinner 3D slice profiles for the gradient sequences and better reconstruction algorithms. These advances may improve the accuracy of both of these tests.
Recently, both ultrasonography10,11 and MRI arthrography12'13 have been proposed to be efficacious for the diagnosis of loose bodies. Further studies are required to validate the effectiveness of these newer methods compared with those which we have studied.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
1. O'Driscoll SW, Morrey BF. Arthroscopy of the elbow: diagnostic and therapeutic benefits and hazards. J Bone Joint Surg [Am] 1992;74-A:84-94.
2. Ogilvie-Harris DJ, Schemitsch E. Arthroscopy of the elbow for the removal of oose bodies. Arthroscopy 1993;9:5-8.
3. Boe S. Arthroscopy of the elbow: diagnosis and extraction of loose bodies. Acta Orthop Scand 1986;57:52-3.
4. Ward WG, Belhobek GH, Anderson TE. Arthroscopic elbow findings: correlation with preoperative radiographie studies. Arthroscopy 1992;8:498-502.
5. Singson RD, Feldman F, Rosenberg ZS. Elbow joint: assessment with double-contrast CT arthrography. Radiology 1986;160:167-73.
6. Murphy BJ. MR imaging of the elbow. Radiology 1992;184:525-9.
7. Quinn SF, Haberman JJ, Fitzgerald SW, et al. Evaluation of loose bodies in the elbow with MR imaging. J Magn Reson Imaging 1994;4:169-72.
8. Shrout PE, Fleiss JL. lntraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420-8.
9. Cohen J. A coefficient of agreement for nominal sales. Educ Psych Meas 1990;20: 37-46.
10. Miller JH, Beggs I. Detection of intraarticular bodies of the elbow with saline arthrosonography. Clin Radiol 2001;56:231-4.
11. Martinoli C, Bianchi S, Giovagnorio F, Pugliese F. Ultrasound of the elbow Skeletal Radiol 2001;3:605-14.
12. Palmer WE. MR arthrography: is it worthwhile? Top Magn Reson Imaging 1996;8: 24-43.
13. Carrino JA, Smith DK, Schweitzer ME. MR arthrography of the elbow and wrist. Semin Musculoskelet Radiol 1998;2:397-414.
J. H. Dubberley,
K. J. Faber,
S. D. Patterson,
J. C. MacDermid,
G. J. W. King
From the University of Western Ontario, London, Canada
* J. H. Dubberley, MD, FRCSC, Orthopaedic Surgeon
St. Boniface General Hospital, University of Manitoba, Z3045-409 Tache Avenue, Winnipeg, Manitoba R2H 2A6, Canada.
* K. J. Faber, MD, FRCSC, Orthopaedic Surgeon
* W. Romano, MD, FRCP, Radiologist
* G. Garvin, MD, FRCP, Radiologist
* J. Bennett, MD, FRCP, Radiologist
* J. C. MacDermid, PhD, Researcher
* G. J. W. King, MD, FRCSC, Orthopaedic Surgeon
The Hand and Upper Limb Centre, St. Joseph's Health Centre, University of Western Ontario, 268 Grosvenor Street, London, Ontario N6A 4L6, Canada.
* S. D. Patterson, MD, FRCSC, Orthopaedic Surgeon
Central Florida Orthopaedic Surgery Associates, PL, 2000 East Edgewood Drive, Lakeland, Florida 33803-3639, USA.
Correspondence should be sent to Dr G. J. W. King; e-mail: email@example.com
©2005 British Editorial Society of Bone and Joint Surgery
doi:10.1302/0301-620X.87B5. 14851 $2.00
J Bone Joint Surg [Br] 2005;87-B:684-6.
Received 7 August 2003; Accepted after revision 16 August 2004
Copyright British Editorial Society of Bone & Joint Surgery May 2005
Provided by ProQuest Information and Learning Company. All rights Reserved