The Spine Journal 10 (2010) 70–72

Commentary

Role of single photon emission computed tomographyin the diagnosis of chronic low back pain

Conor O’Neill, MDa,*, Douglas K. Owens, MD, MSb,c

aCalifornia Spine Diagnostics, 2100 Webster St, # 518, San Francisco, CA 94115, USAbVA Palo Alto Healthcare Systems, 3801 Miranda Ave., Palo Alto, CA 94304-1290, USA

cCenter for Primary Care and Outcomes Research, Center for Health Policy, Stanford University, 117 Encina Commons, Stanford, CA 94305-6019, USA

Received 9 October 2009; revised 27 October 2009; accepted 5 November 2009

COMMENTARY ON: Makki D, Khazim R, Z

DOI of original a

FDA device/drug

Author disclosure

vate investments, Noc

tific advisory board

AspenBio Pharma, we

research support, NIH

NIH, AHRQ).

* Corresponding

St, # 518, San Franc

(415) 600-7835.

E-mail address: c

1529-9430/10/$ – see

doi:10.1016/j.spinee.2

aidan AA, et al. Single photon emission computer-ized tomography (SPECT) scan–positive facet joints and other spinal structures in a hospital-widepopulation with spinal pain. Spine J 2010;10:58–62 (in this issue).

Bone scintigraphy is a method for identifying sites of al-tered metabolic activity because of bone pathology [1]. Themost commonly used agents are diphosphonate derivativesradiolabeled with technetium-99, which have a high affinityfor bone and accumulate in areas of increased osteoblasticactivity and blood flow [2,3]. Single photon emission com-puted tomography (SPECT) is a scintigraphic technique forobtaining multiplanar views, using a rotating gamma cam-era [1]. This improves anatomic localization and is partic-ularly useful in spine imaging, as it allows the activityfrom the vertebral bodies and the individual posterior ele-ments to be viewed separately [1]. The only significant riskrelated to scintigraphy is radiation exposure. The averagedose of radiation associated with a bone scan is 6.3 millisie-verts, which is roughly the same as the dose associated witha lumbar spine computed tomography [4]. It is difficult toquantify the risk this poses for an individual patient, butgiven the rising rates of exposure to ionizing radiation asso-ciated with imaging studies, and the cumulative effects thatexposure has on oncogenesis, this may be a consideration indeciding whether to obtain an SPECT scan [4,5].

rticle: 10.1016/j.spinee.2009.06.004.

status: not applicable.

s: CON (stock ownership, Relievant, Nocimed; pri-

imed; consulting, SpineView, ISTO, Alleeva; scien-

, Relievant); DKO (consulting, GE Healthcare,

llpoint; scientific advisory board, AspenBio Pharma;

, AHRQ; research support, NIH, AHRQ; grants,

author. California Spine Diagnostics, 2100 Webster

isco, CA 94115, USA. Tel.: (415) 600-7830; fax:

oneill@sbcglobal.net (C. O’Neill)

front matter � 2010 Published by Elsevier Inc.

009.11.010

Scintigraphy may allow differentiation of normal, age-related changes from painful lesions in patients withchronic low back pain (CLBP). The relationship betweenthe scintigraphic and radiological appearances of osteoar-thritis is complex [6]. Osteoarthritis results from abnormaljoint loading, which leads to stresses on subchondral boneand secondary bone responses, ultimately resulting inosteophyte formation. Scintigraphy is extremely sensitivein detecting these bone responses, and in fact will show ab-normal uptake before any morphologic changes appear [2].Locally increased bone turnover may persist until osteo-phytes mature and the arthritic joint stabilizes, at whichpoint isotope uptake decreases [2,6,7]. If these boneresponses are related directly (eg, by local production of cy-tokines) or indirectly (eg, as a marker of abnormal loading)to pain then SPECT could be a useful test.

Makki et al.’s aim was ‘‘to evaluate the prevalence ofSPECT scan positive facet joints and other spinal areas,in different age groups in a hospital-wide population withspinal pain,’’ which they did [8]. They found that approxi-mately 40% of SPECT scans indicated facet joint disease,and that the proportion of positive scans increased withage. How to interpret their results is complicated by the dif-ficulty in making a diagnosis of painful facet arthropathy.

A diagnostic test is accurate if it reliably identifies a dis-ease. The best way to determine accuracy is to apply an in-dependent reference standard for a disease to all thepatients who undergo the test. This allows sensitivity, spec-ificity, and likelihood ratio––all familiar measures of accu-racy––to be calculated. However, the usefulness of thesecalculations depends heavily on the validity of the refer-ence standard [9]. For some diseases, such as cancer, thereare reference standards, such as biopsy, that can classify

71C. O’Neill and D.K. Owens / The Spine Journal 10 (2010) 70–72

patients with very low error rates. However, for many, if notmost, diseases reference standards are considerably flawed.In some cases, the reference standard may be no more thana combination of signs and symptoms that experts agreecharacterize the disease (eg, migraines, depression, fibro-myalgia, irritable bowel syndrome). A more robust, but stillimperfect, reference standard is the response to a therapeu-tic trial, such as a bronchodilator to a patient with suspectedasthma. The less perfect the reference standard, the morelikely that measurements of accuracy will be biased [9].

Makki et al. suggest that the appropriate reference stan-dard for facet joint pain is the response to comparative localanesthetic blocks [8]. Some experts agree with this asser-tion [10], but others do not [11,12]. There are several poten-tial sources of error associated with comparative anestheticblocks [12], including placebo responses, aberrant spread oflocal anesthetic [13,14], and anomalous responses becauseof the disrupted pain processing that occurs in patients withchronic pain [15]. Furthermore, there is no strong evidencethat a therapeutic trial of facet rhizotomy in patients witha positive response to comparative local anesthetic blocksrelieves pain. In fact, in the only randomized controlledtrial of facet rhizotomy that used positive comparativeanesthetic blocks as an inclusion criteria the active and con-trol groups ended up with similar outcomes, and the im-provements in pain in the radiofrequency group were ator below the minimal clinically important difference [16].Thus, experts disagree about the validity of comparativelocal anesthetic blocks, the potential for errors has beendemonstrated, and the response to therapy predicated onthe results from them is unconvincing. Given all this, com-parative local anesthetic blocks constitute, at best, a highlyflawed reference standard for painful facet arthropathy.

Even without a valid reference standard, it would be pos-sible to get some evidence on the accuracy of SPECT scans,if the frequency of positive SPECT scans in asymptomaticindividuals (who don’t have back pain and thereforecouldn’t have painful facets) was compared with the fre-quency in patients with CLBP (some of whom could havepainful facets). The groups would have to be matched forthe severity of radiographic spondylosis. If the rates of pos-itive tests were similar in these two groups, then SPECTwould probably not be a very useful test [17]. Makki et al.have given us the frequency of various SPECT findings ina symptomatic population [8], which is a useful start. How-ever, we also need data from a matched asymptomaticpopulation. Collecting these data would probably not beethical, given the radiation exposure involved.

Although it might not be possible to define the accuracy ofSPECT, it would be possible to determine its effect on out-comes, which is actually the most important attribute of a test[17]. After all, there is no point making a diagnosis unless itleads to a better outcome for the patient. The best way to de-termine the effect of SPECTon outcomes would be to do a di-agnostic randomized controlled trial. For example, a group ofpatients with CLBP would undergo SPECT scans and then be

randomized to an intervention, such as a facet rhizotomy. As-suming appropriate methods to control error and bias, includ-ing concealment of the SPECT results from clinicians andpatients, the results from the study could be used to determinethe treatment effect of SPECT [17]. Although laborious andexpensive, data from a trial such as this would provide impor-tant information on the utility of SPECT in CLBP.

One important factor to consider in assessing the valueof SPECT is the potential relationship between scinti-graphic and magnetic resonance imaging (MRI) findings.Most patients will have an MRI before an SPECT scan.Fat-suppressed sequences on MRI readily detect bone mar-row edema, which in adolescents and young adults withpars defects is associated with a positive SPECT scan[18]. Furthermore, soft-tissue changes that accompany facetosteoarthritis––such as synovitis and cartilage degrada-tion––can be detected on MRI and may be related to pain[7,19]. As Makki et al. point out, one MRI finding, synovialmottling on T2-weighted imaging, is highly specific fora positive SPECT [7], which probably eliminates the needfor SPECT in patients with that finding. There may be sim-ilar relationships between SPECT and other MRI findings.

We do not have the data to know the true value of SPECTscans. Observational studies suggest that a positive SPECTscan predicts a positive response to facet corticosteroids[20,21]. It seems unlikely that the expense and radiation ex-posure associated with SPECT justifies using it routinely be-fore a trial of facet corticosteroid injections, which is a lowrisk and relatively inexpensive procedure. There may be cir-cumstances however, where the stakes are high enough towarrant an SPECT scan. For example, an SPECT scan couldbe considered before a facet rhizotomy, especially if the clin-ical findings, imaging results, and/or responses to injectionsare equivocal. A more conventional role for SPECT has beenin surgical planning, particularly in patients with pars defects[22]. However, MRI with the appropriate fat-suppressed se-quences appears to provide equivalent information, at leastin young patients [18]. Ultimately, until we have moredata—which is unlikely to happen in the foreseeable fu-ture––spine physicians and surgeons need to rely on theirknowledge of the basic principles behind bone scintigraphyand the pathophysiology of spinal pain, as well as their clin-ical judgment, in deciding whether and when to order SPECTscans on their patients.

References

[1] Ryan RJ, Gibson T, Fogelman I. The identification of spinal pathol-

ogy in chronic low back pain using single photon emission computed

tomography. Nucl Med Commun 1992;13:497–502.

[2] Omoumi P, Mercier GA, Lecouvet F, et al. CT arthrography, MR

arthrography, PET, and scintigraphy in osteoarthritis. Radiol Clin

North Am 2009;47:595–615.

[3] McKillop JH, Fogelman I. Bone scintigraphy in benign bone disease.

Br Med J (Clin Res Ed) 1984;288:264–6.

[4] Fazel R, Krumholz HM, Wang Y, et al. Exposure to low-dose ioniz-

ing radiation from medical imaging procedures. N Engl J Med

2009;361:849–57.

72 C. O’Neill and D.K. Owens / The Spine Journal 10 (2010) 70–72

[5] Lauer MS. Elements of danger—the case of medical imaging. N Engl

J Med 2009;361:841–3.

[6] Merrick MV. Investigation of joint disease. Eur J Nucl Med 1992;19:

894–901.

[7] Kim KY, Wang MY. Magnetic resonance image-based morphological

predictors of single photon emission computed tomography-positive

facet arthropathy in patients with axial back pain. Neurosurgery

2006;59:147–56.

[8] Makki D, Khazim R, Zaidan AA, et al. Single photon emissioned

computed tomography (SPECT) scan–positive facet joints and 3

other spinal structures in a hospital-wide population with spinal pain.

Spine J 2010;10:58–62.

[9] Reitsma JB, Rutjes AW, Khan KS, et al. A review of solutions for di-

agnostic accuracy studies with an imperfect or missing reference

standard. J Clin Epidemiol 2009;62:797–806.

[10] Manchikanti L, Singh V. Are the results of a multicenter analysis of

radiofrequency denervation success as a function of single diagnostic

block reliable? Spine J 2009;9:704–5.

[11] Cohen SP, Stojanovic MP. Treatise on facet joint pain, King Midas,

and existensialism. Spine J 2009;9:705–6.

[12] O’Neill C, Owens DK. Lumbar facet joint pain: time to hit the reset

button. Spine J 2009;9:619–22.

[13] Cohen SP, Raja SN. Pathogenesis, diagnosis, and treatment of lumbar

zygapophysial (facet) joint pain. Anesthesiology 2007;106:591–614.

[14] Dreyfuss P, Schwarzer AC, Lau P, Bogduk N. Specificity of lumbar

medial branch and L5 dorsal ramus blocks. A computed tomography

study. Spine 1997;22:895–902.

[15] North RB, Kidd DH, Zahurak M, Piantadosi S. Specificity of diag-

nostic nerve blocks: a prospective, randomized study of sciatica

due to lumbosacral spine disease. Pain 1996;65:77–85.

[16] Nath S, Nath CA, Pettersson K. Percutaneous lumbar zygapophysial

(Facet) joint neurotomy using radiofrequency current, in the manage-

ment of chronic low back pain: a randomized double-blind trial.

Spine 2008;33:1291–7; discussion 1298.

[17] Knottnerus JA, Buntinx F, editors. The evidence base of clinical

diagnosis: theory and methods of diagnostic research (evidence-based

medicine). 2nd ed. London: BMJ Books, 2008.

[18] Campbell RS, Grainger AJ, Hide IG, et al. Juvenile spondylolysis:

a comparative analysis of CT, SPECT and MRI. Skeletal Radiol

2005;34:63–73.

[19] Czervionke LF, Fenton DS. Fat-saturated MR imaging in the detec-

tion of inflammatory facet arthropathy (facet synovitis) in the lumbar

spine. Pain Med 2008;9:400–6.

[20] Dolan AL, Ryan PJ, Arden NK, et al. The value of SPECT scans in

identifying back pain likely to benefit from facet joint injection.

Br J Rheumatol 1996;35:1269–73.

[21] Ackerman WE 3rd, Ahmad M. Pain relief with intraarticular or me-

dial branch nerve blocks in patients with positive lumbar facet joint

SPECT imaging: a 12-week outcome study. South Med J 2008;101:

931–4.

[22] Gregory PL, Batt ME, Kerslake RW, et al. The value of combining

single photon emission computerised tomography and computerised

tomography in the investigation of spondylolysis. Eur Spine J

2004;13:503–9.