Diagnostic imaging for spinal disorders in the elderly: a narrative review

Radiographs are not routinely indicated in the following settings:

Radiographs are indicated in the following settings:

CT or MRI should also be considered in the above settings. Nuclear medicine (bone) scan may be helpful when radiographs are normal or equivocal for fracture.

Reinus and colleagues studied indications for lumbosacral spine radiographs in 482 patients presenting to a Level II emergency department. The major indications for lumbosacral radiographs were lower back pain (92 %) and trauma (36 %). However, patient expectation and medicolegal concerns, related either to insurance documentation or to physician litigation, were cited in 42 % of cases despite the fact that these are not appropriate indications for imaging. They concluded that their data supported the use of lumbosacral spine radiographs for patients with a history of trauma, even if relatively minor, in elderly patients and in patients with lower back pain who have a history of neoplasm. However, the data revealed that lumbosacral radiographs obtained for an isolated complaint of lower back pain or isolated neurologic abnormalities generally provide no clinically useful information. They concluded that such patients are better examined (although not necessarily at the time of emergency department evaluation) with techniques such as MR imaging that reveal soft-tissue lesions. [70]

In alert and stable cervical spine trauma patients, radiographs are only routinely indicated in patients with positive high-risk factors on the Canadian Cervical Spine Rule for Radiography in Alert and Stable Trauma Patients (CCSR) [66, 71]. One of those factors is age over 65. Therefore, all patients over age 65 should get a 3-view routine cervical spine radiographic series (anterior-posterior, lateral, and anterior-posterior open mouth), in acute cervical spine trauma. If fracture is suspected, CT is recommended rather than oblique, pillar or flexion-extension radiographs. MRI may also be indicated in certain cases to evaluate soft tissue, cord or nerve root injury. (Figure 1)

The Bone and Joint Decade 2000–2010 task force on neck pain and its associated disorders (TFNP) concluded that CT scans have better validity (in adults and elderly) than radiographs in assessing high-risk and/or multi-injured blunt trauma neck patients. There is no evidence, on the other hand, that specific MRI findings are associated with neck pain, cervicogenic headache, or whiplash exposure. Furthermore, flexion-extension radiographs and 5-view radiographs (cross table lateral, anterior-posterior, bilateral oblique, and odontoid views) in the acute stage of blunt neck trauma add little to static radiography in predictability and accuracy [31].

Conventional radiographs are not initially indicated in adult patients with acute, subacute, or persistent uncomplicated LBP with no neurologic deficits or red flags. As a general rule, a 4–6 week therapeutic trial of conservative care is appropriate before radiographs are obtained. However, since age 65 or over is considered a red flag, radiographs are often indicated at the time of initial presentation, especially if the patient has at least one additional red flag. Additionally, lumbar spine radiographs are indicated in patients over 65 or those who have progressive neurologic deficits with suspected degenerative spondylolisthesis, lateral stenosis, or central stenosis. Oblique or flexion-extension radiographs, CT or MRI are not initially indicated in these patients and should be reserved for those with a failed 4–6 week trial of conservative care or deteriorating neurologic deficit or disabling leg pain.

According to the North American Spine Society (NASS) evidence-based clinical guideline for the diagnosis and treatment of DLSS, MRI is the most appropriate, noninvasive test for imaging degenerative lumbar spinal stenosis. These guidelines further recommend that CT myelography is useful in patients who have contraindications to MRI, patients with MRI findings that are inconclusive, or patients with a poor correlation between symptoms and MRI findings. CT without myelography is useful in patients who have contraindications to MRI, patients with MRI findings that are inconclusive, or patients with a poor correlation between symptoms and MRI findings and those who are not candidates for CT myelography. [72]

According to the Evidence-Based Clinical Guideline developed by the DLS Work Group of NASS, the most appropriate, noninvasive test for detecting DLS is the lateral radiograph, whereas the most appropriate, noninvasive test for imaging the stenosis associated with DLS is MRI. (Figure 2) As in imaging recommendations for DLSS, plain myelography or CT myelography are also useful for assessing spinal stenosis associated with DLS. CT without myelography is a useful noninvasive study in patients who have contraindications to MRI, patients with MRI findings that are inconclusive, or patients with a poor correlation between symptoms and MRI findings and those who are not candidates for CT myelography [73].

Conventional radiographs are not initially indicated in suspected acute lumbar disc herniation (protrusion, extrusion, sequestration) unless the patient is over age 50 or has progressive neurologic deficits. However, radiographs are insensitive to disc herniations and acute disc herniations occur mostly in the 35–54 year age range. While degenerative disc bulges are more likely to occur in older individuals, they are not visible on radiographs either [52].

One of the difficulties in evaluating the utility and validity of MRI in LBP is the high prevalence of abnormal findings in asymptomatic individuals. A recent systematic review and meta-analysis by Endean et al concluded that MRI findings of disc protrusion, nerve root displacement or compression, disc degeneration, and high intensity zone are all associated with LBP, but that individually, none of these abnormalities provides a strong indication that LBP is attributable to underlying pathology [74]. (Figure 3) This limits the value of abnormal MR imaging findings in evaluating intervertebral disc disorders and degenerative changes in elderly patients with LBP.

Kalichman et al retrospectively evaluated spinal degeneration in a subset of 187 participants with a mean age of 52.6 years of age who initially underwent multidetector CT scans primarily to assess aortic calcifications. While degenerative changes were extremely prevalent, the only degenerative feature associated with self-reported LBP was spinal stenosis. Intervertebral disc space narrowing (present in 63.9 % of spines), and facet joint osteoarthrosis (64.5 %) were unassociated with LBP [48].

Conventional radiographs or special investigations are not initially indicated in uncomplicated (no neurologic deficits or red flags), nontraumatic neck pain of less than four weeks duration. Radiographs are indicated, however, for patients with nontraumatic neck pain and radicular symptoms. This category includes patients with suspected acute cervical disc herniation or suspected acute cervical spondylotic radiculopathy or lateral canal stenosis. While the three-view series of radiographs are suggested, oblique or swimmer (spot lateral cervicothoracic) views may also be included. Cervical spine MRI should be considered after a failed four-week trial of conservative therapy.

The TFNP recommends that radiographs are not even initially indicated in patients with uncomplicated subacute (4–12 week duration) and persistent (>12 week duration) neck pain with or without associated arm pain. They recommend a system of “Red Flags” (similar to those now used in assessing patients with low back pain), that allow clinicians to rule out serious pathology in patients seeking care for neck pain with no exposure to blunt trauma. (Table 2) Important serious disorders to consider include pathologic fractures, neoplasm, systemic inflammatory disease, infection, cervical myelopathy, and/or previous cervical spine or neck surgery or open injury [31].

Elderly patients with CES (presenting as LBP, bilateral or unilateral sciatica, saddle anesthesia, motor weakness of the lower extremities that may progress to paraplegia, urinary retention, or bowel and/or bladder incontinence) should be treated as a surgical emergency requiring immediate emergency referral. There is no value in obtaining imaging prior to the referral as the imaging studies will likely be repeated at the emergency facility [65].

AAA is a vascular disease with life-threatening implications that affects about 4-9 % of men and 1 % of women, mostly age 65 and over. AAA commonly presents as back pain and therefore may be encountered in elderly patients seeking chiropractic care. In non dissecting AAAs, medical referral and ultrasound are recommended even if conventional radiographs are negative (calcification, the most reliable radiological sign, is seen in only 50 % of AAA) [75]. In 2005 the US Preventive Services Task Force (USPSTF) published a recommendation that all men between the ages of 65 and 75 who are or have been smokers should have a one-time abdominal diagnostic ultrasound study (DUS) to screen for AAA. They emphasized that 70 percent of men in this age group have smoked and would benefit from routine screening to check for aneurysms. The USPSTF make no recommendation about AAA screening for men between the ages of 65 and 75 who have never been smokers and they recommend against such routine DUS screening for AAA in women [76]. In the US, Medicare covers the cost of this one-time screening DUS in patients with a family history of AAA or who have smoked at least 100 cigarettes in their lifetime [77]. (Figure 4)

Suspected acute AAA or thoracic aortic aneurysm, dissection, rupture, occlusion or traumatic injury in any patient requires immediate emergency referral without imaging [52].

Conventional radiographs are notoriously unreliable for assessing bone mineral density (BMD). In elderly patients with or without fragility fractures, dual energy x-ray absorptiometry (DXA) is indicated to detect and quantify osteoporosis. The decision to test BMD should be based on a woman's clinical risk profile, as well as the potential impact of results on management [78]. Regardless of clinical factors, all women over age 65 and all males over age 70 should be tested for BMD. BMD testing is also recommended for postmenopausal women younger than 65 with osteoporotic risk factors and in men aged 50–69 if at least one major or two minor risk factors for osteoporosis are present [78]. Several of these important osteoporosis risk factors have been identified that place elderly patients, especially postmenopausal females, at risk. (Table 3) The FRAX® tool was developed by the World Health Organization to evaluate fracture risk in both postmenopausal women and men aged 40 to 90 years. It is validated to be used in untreated patients only. The current National Osteoporosis Foundation Guide is based on individual patient models that integrate the risks associated with clinical risk factors as well as BMD at the femoral neck. The FRAX® algorithms give the 10-year probability of fracture of the spine, forearm, femoral neck, or proximal humerus [79]. Simplified paper versions, based on the number of risk factors can be downloaded for office use at: http://​www.​shef.​ac.​uk/​FRAX/​. For most people, an interval of at least two years is an appropriate duration for repeating BMD testing.

In the US, Medicare covers the cost of DXA scans once every 24 months to determine fracture risk in people who are at risk for osteoporosis [80]. In Australia, Medicare has covered bone mineral density tests for all patients aged 70 years and over since April 2007 [81]. (Figure 5)

Conventional radiographs are indicated for the initial evaluation of suspected thoracic and lumbar spine compression fractures. Additional MRI or CT evaluation is indicated in cases where initial radiographs are positive, difficult to interpret, or when complex lesions or ligamentous instability or neural injuries are suspected. (Figure 2) MRI is also useful in determining whether fractures are acute or chronic and also prior to kyphoplasty procedures, for surgical planning, and to detect incidental pathology [82]. Fluorodeoxyglucose positron emission tomography fused with computed tomography (FDG-PET/CT) is useful in differentiating benign from malignant compression fractures [83]. The use of PET/CT is limited, however, by its considerable expense.

Myeloma is the most common primary malignant bone tumor and accounts for about 10 % of all hematologic malignancies [84]. Three diagnostic critieria must be present: (a) greater than 10 % atypical marrow plasma cells and/or biopsy-proved plasmacytoma; (b) monoclonal paraprotein, and (c) myeloma-related organ dysfunction. A bone marrow biopsy or aspirates are necessary to confirm the diagnosis [85]. Myeloma typically infiltrates active red marrow tissue and destroys bone. Typical sites of involvement include the skull, spine, pelvis, ribs, humerus and femur. Initially, radiographs frequently appear normal. Later on, osteoclast stimulation and osteoblast suppression result in diffuse osteopenia which may be difficult to differentiate from senile osteoporosis. With further disease progression, multiple well-circumscribed radiolucencies predominate. Multislice helical axial CT with coronal and sagittal reconstructions is more sensitive than radiographs. (Figure 6) Osteoblastic lesions are extremely rare in myeloma. MRI is more useful than radiography or CT for staging the disease and detecting various patterns of marrow infiltration. FDG-PET can be used to detect multiple myeloma with good sensitivity and specificity. Its ability to assess metabolic activity can be useful, especially when evaluating treatment response and monitoring relapse [84]. Additionally, FDG-PET has been shown superior to conventional radiography but less so compared with MRI [86]. In myeloma patients, NM bone scans may show photopenic areas or a negative scan resulting in false negative interpretations [87].

Metastasis of cancer to the bones is the most common malignant process of the skeleton. More than 80 % of adult cases originate from primary carcinoma of the prostate, breast, lung and bronchus, thyroid, and kidney. Skeletal metastasis is 25 to 30 times more common than any primary bone tumor and as many as 140,000 new cases are identified in the United States annually. Most cases result in osteolytic bone destruction, but some cases are purely osteoblastic or a combination of osteolytic/osteoblastic involvement [84]. The ideal imaging technique for initial staging and monitoring should quickly and accurately identify all active sites of the disease, but no single imaging modality satisfies all the criteria in different situations. MRI, CT, NM bone scan, FDG-PET, and PET/CT are all useful, and any of these may be the best study for an individual patient, depending on their unique clinical circumstances [84]. (Figure 7)

In 2009, the American College of Radiology (ACR) updated their appropriateness criteria for imaging of metastatic bone disease. The ACR reviewed published meta-analyses and systematic reviews with evidence tables focusing on the utility of imaging examinations in differential diagnosis. In summary, they concluded that NM bone scanning is the most widely used primary imaging examination for detecting osseous metastasis. NM is sensitive in detecting osseous abnormalities, but it is nonspecific. Therefore, after an abnormality has been detected, radiographs should be obtained to make sure the abnormality does not represent a benign process. If radiography is not diagnostic, additional lesion workup with MRI, CT, SPECT, or FDG-PET/CT is highly variable and should be based on the clinical situation and lesion location [87]. (Figure 7)

Conventional radiography has low sensitivity for bone destruction and may result in false negative interpretations in cases of skeletal metastasis. It is for this reason that some current guidelines emphasize the use of advance imaging of the spine instead of radiographs to make or to exclude the diagnosis of spinal metastases [88]. Medical referral is recommended if primary contact practitioners do not have immediate access to advanced imaging techniques such as MR imaging, CT, and NM bone scanning studies. In geographical areas with limited access to primary care physicians, it some authorities argue that it is reasonable to include conventional radiography (with or without erythrocyte sedimentation rate) in the evaluation of persons with red flag indicators of suspected skeletal metastasis [55].

MSCC represents compression of the dural sac and its contents—the spinal cord and cauda equina—by an extradural mass. Metastatic lung, breast and prostate cancers are the commonest malignancies causing MSCC and account for over 50 % of cases [89]. In 7 % of patients the site of primary tumour may remain unidentified [90]. In 23 % of patients, MSCC will be the first presenting problem. Because patients with known malignancy may also have spinal cord compression from a non-malignant cause, it is important to differentiate MSCC from other causes such as degenerative stenosis and osteoporotic compression fractures [91]. According to recommendations in 2008 by the National Collaborating Center for Cancer, every cancer network should ensure that there is local access to urgent magnetic resonance imaging (MRI) within 24 hours for all patients with suspected MSCC. This service should be available outside normal working hours and with 24-hour capability in centres treating patients with MSCC [91]. More specific Selected Imaging Recommendations from 2008 NICE Guidelines on MSCC [88] are as follows.

MRI of the spine in patients with suspected MSCC should be supervised and reported by a radiologist and should include sagittal T1 and/or short T1 inversion recovery (STIR) sequences of the whole spine, to prove or exclude the presence of spinal metastases. Sagittal T2 weighted sequences should also be performed to show the level and degree of compression of the cord or cauda equina by a soft tissue mass and to detect lesions within the cord itself. Supplementary axial imaging should be performed through any significant abnormality noted on the sagittal scan.

Consider targeted CT scan with three-plane reconstruction to assess spinal stability and plan vertebroplasty, kyphoplasty or spinal surgery in patients with MSCC.

Consider myelography if other imaging modalities are contraindicated or inadequate.

Myelography should only be undertaken at a neuroscience or spinal surgical centre because of the technical expertise required and because patients with MSCC may deteriorate following myelography and require urgent decompression.

Do not perform plain radiographs of the spine either to make or to exclude the diagnosis of spinal metastases or MSCC.

In patients with a previous diagnosis of malignancy, routine imaging of the spine is not recommended if they are asymptomatic. (Serial imaging of the spine in asymptomatic patients with cancer who are at high risk of developing spinal metastases should only be performed as part of a randomised controlled trial.)

Perform MRI of the whole spine in patients with suspected MSCC, unless there is a specific contraindication. This should be done in time to allow definitive treatment to be planned within 1 week of the suspected diagnosis in the case of spinal pain suggestive of spinal metastases, and within 24 hours in the case of spinal pain suggestive of spinal metastases and neurological symptoms or signs suggestive of MSCC, and occasionally sooner if there is a pressing clinical need for emergency surgery.

Causes of cord compression include trauma, tumors, infection, vascular disease, degenerative conditions, demyelinating disorders, spinal stenosis, & central cervical disc herniation [52]. Clinicians should be aware that nearly all of the clinical tests for cervical spine myelopathy are poor screening tools, which implies that manually oriented clinicians may mistakenly proceed with treatment when it is not indicated. [92]. Patients presenting with signs and symptoms of cervical spine myelopathy should therefore undergo appropriate investigation before proceeding with manual therapy interventions. Cervical spine radiographs including oblique projections are indicated in patients with suspected cervical compressive myelopathy or radiculo-myelopathy. MRI should also be performed to identify cord compression and/or high signal intensity intramedullary cord lesions, the latter of which are associated with a poorer prognosis even after decompressive surgery. If MRI is unavailable, CT-myelography should be considered. In addition to imaging, electrophysiologic testing such as somatosensory evoked potentials (SSEP) may be useful [52, 93].

AAI is of particular importance to chiropractors and other clinicians involved in manual therapy of the cervical spine. Many conditions result in osseous abnormalities such as nonunion or agenesis of the odontoid, rupture, laxity, or absence of the transverse ligament, or other upper cervical spine pathologies. These include but are not limited to: a) active inflammatory arthritis such as rheumatoid arthritis (RA), psoriatic arthropathy, ankylosing spondylitis, and systemic lupus erythematosus; b) Congenital disorders and hereditary connective tissues disorders such as spondyloepiphyseal dysplasia, os odontoideum, and several syndromes including Klippel-Feil, Morquio, Down (20 % of Trisomy 21 patients are born without a transverse ligament), Ehlers-Danlos type III, and Marfan; c) traumatic conditions such as C1 or C2 fracture or dislocation. Lateral radiographs of the cervical spine obtained in flexion and extension are indicated in suspected AAI, however, a single lateral cervical radiograph with the patient in supervised comfortable flexion should reveal any subluxation in patients with suspected instability. In adults, the atlantodental interval should not exceed 3 mm in neutral, flexion, or extension neck positions. In the presence of neurologic signs and symptoms, MRI or CT are indicated to reveal osseous abnormalities, stenosis, and spinal cord lesions [52, 94].