Richard Bartley MSc MCSP, Betsi Cadwaladr University Health Board, Wales
Magnetic resonance imaging (MRI) has, over the last twenty years, become the imaging technique of choice for many clinicians. Because MRI has excellent contrast sensitivity it is capable of producing detailed images of numerous morphological features in the spine, including the presence of a disc herniation and root compression (Steinmetz, 1987).
MR images may be obtained in sagittal, axial and coronal planes and simultaneous acquisition of multiple slices makes the modality cost-effective (Crooks et al, 1983).
No harmful effects of MR imaging to patients has been identified. The properties of a specific imaging system are measured in terms of spacing resolution (i.e. the thickness of the 'slices' being observed) and the contrast and sharpness of the images (Margulis & Demas, 1985). The greater these properties the more accurate reporting can become, particularly in detecting subtle differ-ences in discogenic disease and nerve root compression.
MRI works by placing patients in a static external magnetic field and then subjecting them to repeated short bursts of radio waves of a specific radio frequency (RF). Protons are excited and then relaxed which release energy. This release of energy produces signals which can be decoded into an image (Steinmetz, 1987).
As with all high technological advances, MRI has its own terminology. The repetition time (TR) represents the time between RF pulses. The echo time (TE) represents the time between the application of the RF pulse and the time of recording the MR signal. These parameters can be adjusted by the radiologist to produce two images:
(i) T1 weighted images (which reflects the time required for realignment of the protons within the magnetic field) are excellent for identifying structures containing fat, haemorrhage or proteinaceous fluid and are produced by setting a long TR and a short TE. Disc disease and herniation are well observed on T1 images.
(ii) T2 weighted images (which reflects the time the protons take to lose coherence following excitation) show structures rich in extra cellular water, such as cerebro-spinal fluid, neoplasm, discs, and are produced by setting a long TR and a long TE.
These parameters can be manipulated to produce images that emphasise the differences between diseased tissues and normal tissue. Careful selection of these pulse sequences is critical to diag-nostic utility (Margulis & Demas 1985). Their use enables a number of morphological structures to be identified.
Reduction in lumbar disc height
Dehydration reduces the sagittal height of the lumbar disc and possibly the lateral recesses in which the nerve roots lie. The change in height may be a important contributing factor in root compression (McNally, 1997).
Disc herniation
Disc hernias have numerous properties which can be observed with MRI:
(i) Position - The hernia can be in a central, paracentral or lateral position. Edgar and Parkes, 1974, observed that the paracentral position was the most critical in spinal nerve root compression.
(ii) Size - This is usually calculated by measuring the diameter of the central canal
(iii) Shape - The hernia can be narrow or wide. A narrow, pointed hernia may be more likely to cause complete compression (Thelander, 1991).
Root compression due to a paracentral disc hernia
Which may be complete, partial (displaced) or absent.
Other manifestations
Facet joint disease, pars defects and spondylolysthesis, central and lateral canal stenosis (identified by measuring the diameter of the spinal canal), tumours and inflammatory disease.
The use of gadolinium enhancement can further highlight phenomena such as tumours, infections and nerve root scarring.
Reliability of MRI
Reliability is important to the integrity of any diagnostic tool as it enables clinicians to make con-sistent comparisons (LoBiondo-Wood & Haber, 1990). Modic et al (1986) prospectively studied 60 patients with suspected lumbar prolapsed disc and/or canal stenosis with MRI and myelography (the use of a radiopaque contrast medium used with plain x-ray) and compared the result with the results of surgery (the 'gold standard'). The overall accuracy was 82.6% for MRI compared with 71.8% for myelography. Jackson et al (1988) carried out a similar study and found the false-positive rate (i.e. discs which appear prolapsed on MR imaging but found not to be during surgery) was only 13% compared with myelography at 21%. However it is not clear from these studies whether there was significant root compression present.
Interpretation is equally as important as technical capacity. Unreliable interpretation of an imaging modality by investigators with different levels of skill result in performance bias (Boos & Lander, 1996). Raininko et al (1995) investigated the reliability of accurately interpreting MRI scans looking specifically at identifying signs of disc degeneration. The authors found acceptable levels of intra-observer reliability but considerable variability in standards of inter-observer reliability.
Herzog (1993) points out that bias can occur because observers may classify their findings differently however accurate they may be. This is well exemplified by the Brant-Zawadzki et al (1995) study on inter- and intra-observer variability in interpreting lumbar disc abnormalities. Agreement was generally high between observers and amongst the observers themselves (80% and 86% respectively). However the main area of disagreement was between normal and bulging discs. The authors concluded that this was due to differences in grading systems rather than inaccurate readings.
Axial MRI (T2) view of lumbar segment L4/5
Showing a paracentral herniation - seen in the right inferior angle of the disc.
Validity of MRI
Magnetic resonance imaging has become an important tool in the diagnosis of sciatica. However surgical management based purely on MRI findings is inappropriate unless they are matched to clinical findings (Boden et al, 1990). Validity is crucial to the role of any diagnostic tool, as the instrument must be seen to be measuring what it is intending to measure (LoBiondo-Wood & Haber, 1990). However much controversy remains over the validity of MRI findings in relation to clinical conditions such as low back pain and sciatica. Mulholland & McCall (1992) state that whilst MRI findings may have great importance in the clinical problem of one patient, it may have no significance in the condition of another. Boos, 1995, tells us that the presence of a disc hernia on an MRI scan per se may not necessarily be a cause of radicular pain due to nerve root compression.
Numerous papers have highlighted the incidence of abnormal morphological findings in asympto-matic volunteers (Weinreb 1989, Boden 1990, Tertti 1991, Buirski 1993, Powell 1995, Boos 1996). These are termed 'false positives'. Boden et al in 1990 performed MRIs on 67 individuals who had never had low back pain or radicular pain. Of those under 60 years of age 20% had a disc hernia whilst the figure was higher at 36% in the over 60 year old. Jensen et al (1994) carried out a similar study on 98 asymptomatic people and found that only 36% had normal discs at all levels, 52% had a minor disc hernia at each level and 27% had a significant disc hernia. They concluded that the presence of disc hernias may not be clinically relevant.
A possible flaw with such studies is the potential failure to match patients by age, sex and occup-ational risk factors (sedentary/manual job). Boos et al (1996) determined the incidence of disc hernia in a matched group of 46 asymptomatic volunteers and assessed the diagnostic accuracy of MRI in identifying symptomatic disc hernias. Both groups consisted of 46 volunteers which were matched for age, sex and work status. All patients underwent an MRI with the results recorded by two 'blind' independent radiologists.
The results showed almost equal amount of disc degeneration in both groups and a high incidence of disc hernia in the asymptomatic patients (76% compared with 96% in the symptomatic group). However an important finding was that neural compromise (root compression) was highly significant (<0.0001) in the symptomatic group. This study demonstrated that MRI is able to identify root compression with some accuracy.
The presence of findings in symptomatic patients also requires careful scrutiny to ensure that they are valid and sensitive to the clinical features and degree of disability presented to the clinician. In other words, the MRI findings must correspond to the clinical symptoms and signs and level of disability reported, if they are to have any role in the clinical decision making process.
References
Brant-Zawadzki MN, Jensen MC, Obuchowski N, Ross JS, Modic MT: Interobserver and intraobserver variability in interpretation of lumbar disc abnormalities.Spine 1995; 11: 1257-1264
Boden SC, Davis DO, Dina TS, Patronas NJ, Wiesel SW: Abnormal magnetic-resonance scans of the lumbar spine in asymptomatic subjects. Bone joint Surg [Am] 1990; 72: 403-408
Boos N, Lander PH: Clinical efficacy of imaging modalities in the diagnosis of low back pain disorders.Eur Spine J 1996; 5: 2-22
Buirski G, Silberstein M: The symptomatic lumbar disc in patients with low back pain: magnetic resonace imaging appearances in both a symptomatic and control population.Spine 1993; 18: 1808-1811
Crooks LE: Clinical efficacy of nuclear Magnetic Resonance ImagingRadiology 1983; 146: 123-126
Edgar MA, Park WM: Induced patterns on passive striate leg raising in lower lumbar disc protrusion: a clinical, myelographic and operative study in fifty patients.J Bone Joint Surg[Br] 1974; 56: 658-667
Herzog RJ: Technology assessment: a case study in diagnostic imaging.American Academy of Orthopaedic Surgeons - Knowledge update 4, Illinois 1993
Jackson RP, Cain JE, Jacobs PR, Cooper BR, McManus GE: Neuroradiographic diagnosis of lumbar herniated nucleus pulposus with CT, myeolography, CT-myelography and magnetic resonance imaging.Spine 1989, 14: 1362-1367
Jenson MC, Brant ZM, Obuchowski N, Modic MT, Malaksian D, Ross JS: Magnetic resonance imaging of the lumbar spine in people without back pain.N Engl J Med 1994; 331: 69-73
LoBiond-Wood G, Haber J: Nursing Research: Methods, Critical Appraisal and Utilisation.Mosby Publications, St Louis, 1990
Margulis AR, Demas B: The Current Status of Magnetic Resonance Imaging in Medical Diagnosis.Contemporary Imaging, Ed. Goldberg HI, Higgins CB & Ring EJ, University of California, 1995
McNally E: Unpublished report. Nuffield Orthopaedic Centre, Oxford
Modic MT, Marsaryk T, Bumphrey F, Goormastic M, Bell G: Lumbar herniated disc disease and canal stenosis: prospective evaluation by surface coil.AJR 1986; 147: 757-765
Mulholland R, McCall I: Magnetic Resonance Imaging of the Lumbar Spine.The Lumbar Spine - The International Society for the Study of the Lumbar Spine, Saunders, Phil. 1995
Powell MC, Wilson M, Szpryt P, Symond EM, Worthington BS: Prevalence of lumbar disc degeneration observed by magnetic resonance imaging in symptomless women.Lancet 1986; 2: 1366-1367
Raininko R, Manninen H, Battie MC, Gibbons LE, Gill K, Fisher LD: Observer variability in the assessment of disc degeneration on magnetic resonance images of the lumbar and thoracic spine.Spine 1995; 20: 1029-1035
Tertti MO, Salminen JJ, Paajanen HE, Terho PH, Kormano MJ: Low back pain and disk degeneration in children: a case control MR imaging study.Radiology 1991; 180: 503-507
Weinreb JC, Wolbarsht LB, Cohen JM, Brown CEL, Maravilla KR: Prevalence of lumbosacral intervertebral disk abnormalities on images in pregnant and asymptomatic non-pregnant women.Radiology 1989; 17: 125-1