Associations Between Serum Biomarkers and Pain and Pain-Related Function in Older Adults with Low Back Pain: A Pilot Study Gwendolyn A. Sowa, MD, PhD,a,b Subashan Perera, PhD,c Bernard Bechara, MS,b Vikas Agarwal, MD,d John Boardman, MD,d Wan Huang, MD,a,b Alejandra Camacho-Soto, MD,a,b Nam Vo, PhD,b James Kang, MD,b and Debra Weiner, MDc,e,f,g,h
OBJECTIVES: To examine the relationship between serum biomarkers and self-reported pain intensity and pain-related function, in addition to the contribution of magnetic resonance imaging (MRI) findings of lumbar spine degenerative changes, in older adults with chronic low back pain. DESIGN: Single-center cross-sectional cohort study. SETTING: Academic medical center. PARTICIPANTS: Individuals aged 60 and older with axial low back pain without radiculopathy or previously diagnosed osteoarthritis of the knee or hip or pain outside the low back that is more severe than the back pain (n = 43). MEASUREMENTS: To examine pain-related impairment, pain was measured on a pain thermometer and the McGill Pain Questionnaire Short Form was administered. To examine pain-related function or activity limitation, the Roland Morris Disability Questionnaire, Short Physical Performance Battery (SPPB), and repetitive trunk rotation were used. Single plasma samples were obtained before and after physical performance tests and analyzed for inflammatory markers (E-selectin and regulated on activation, normal T cell expressed and secreted (RANTES)), inhibitors of catabolic enzymes (tissue inhibitor of metalloproteinases-1 (TIMP-1)), markers of matrix turnover (C- telopeptide of type II collagen (CTX-II) and aggrecan chondroitin sulfate 846 (CS846)), and stress biomarkers (neuropeptide Y From the aDepartment of Physical Medicine and Rehabilitation; bFerguson Laboratory for Orthopaedic Research, Department of Orthopaedics; c Department of Medicine; dDepartment of Radiology; eDepartment of Psychiatry; fDepartment of Anesthesiology; gClinical and Translational Sciences Institute, University of Pittsburgh; and hGeriatric Research, Education and Clinical Center, Veterans Affairs Pittsburgh Healthcare System, Pittsburgh, Pennsylvania. Address correspondence to Gwendolyn Sowa, Associate Professor, Department of Physical Medicine and Rehabilitation, Co-Director, Ferguson Laboratory for Orthopaedic and Spine Research, University of Pittsburgh, 3471 5th Ave, Suite 201, Pittsburgh, PA 15213. E-mail: [email protected] DOI: 10.1111/jgs.13102
JAGS 62:2047–2055, 2014 © 2014, Copyright the Authors Journal compilation © 2014, The American Geriatrics Society
(NPY)). Conventional nongadolinium lumbar MRI was performed and analyzed quantitatively and clinically. RESULTS: Composite MRI measurements did not show significant correlation with pain or pain-related function. Basal levels and changes in serum biomarkers in response to activity, particularly NPY and RANTES, demonstrated associations with pain and pain-related function in addition to the explanatory power of MRI-based results. CONCLUSION: Serum biomarkers may be a metric for assessment of active disease in older adults, in whom imaging changes are ubiquitous. In addition, changing levels of biomarkers in response to activity suggests that they may be useful as metrics to measure treatment responses in future studies and may reflect potential targets for use in designing personalized treatment for older adults with low back pain. J Am Geriatr Soc 62:2047–2055, 2014.
Key words: low back pain; biomarkers; imaging
U
nderstanding the contribution of back pain to the function of older adults is of paramount concern because of the increasing prevalence of low back pain with age and expectations of older adults to remain active, but current tools to assess lumbar spine pathology focus on imaging findings and do not address pain or catabolic activity of the tissue. Although intervertebral disc degeneration is ubiquitous on magnetic resonance imaging (MRI) of older adults, only 42% report low back pain during the prior 12 months.1 The lack of correlation between anatomical changes and symptoms in older adults2,3 makes imaging a suboptimal predictive tool in identifying pain generators in older adults with low back pain. Despite the limitations of current imaging modalities, correct identification of pain generators is critical to guiding appropriate and successful treatments. This is of paramount concern in older adults, in whom multiple pain generators frequently coexist. Nevertheless, interventions
0002-8614/14/$15.00
2048
SOWA ET AL.
are frequently planned based on imaging studies, often resulting in unnecessary procedures and associated morbidity.4 Because serum levels of inflammatory and matrix breakdown products address disease activity in other musculoskeletal conditions,5,6 evaluation of biomarkers for low back pain has the potential to provide assessment of active pain generators in addition to the changes observable on imaging studies. Ascertaining which of these molecular pathways are linked to pain and pain-related functional limitations may differentiate individuals requiring therapeutic intervention from those in whom identification of alternative pain generators should be explored. In addition, evaluation of these markers has the potential to contribute to the understanding of the mechanism of the degenerative process and low back pain and may therefore lead to novel diagnostic and therapeutic tools. Candidate biomarkers chosen for this pilot study are important mediators in the inflammatory, catabolic, and pain pathways. Biological mediators of inflammation and catabolism are involved in intervertebral disc degeneration.7–10 In other degenerative conditions, inflammatory and catabolic biomarkers are sensitive indicators of disease activity and progression,5,6 and high levels of inflammatory cytokines and catabolic mediators have been isolated from degenerative discs.8,11,12 There are dramatic differences between the expression profiles of inflammatory and catabolic genes in an animal model of pathologic disc degeneration and a normal aging animal, with differences noted in critical inflammatory and catabolic genes.13 E-selectin and regulated on activation, normal T cell expressed and secreted (RANTES) were chosen as markers of inflammation. In addition to E-selectin’s known role as a marker of systemic inflammation, high levels are associated with positive straight leg raise in individuals with surgically confirmed nerve root compression14 and are a useful serum prognostic indicator in rheumatoid arthritis.15 More relevant to the current study, positive immunostaining for E-selectin has been found in surgically obtained herniated disc specimens, although all of these subjects had associated radicular pain.16 Nevertheless, the presence of positive staining in the disc suggests a possible role in axial low back pain as well. RANTES is expressed in response to inflammatory stimuli and affects catabolic activity by altering matrix metalloprotease production in articular cartilage.17 Human intervertebral discs also produce RANTES,18 serum levels have been shown to change in response to activity in other disease conditions,19 and painful discs express higher levels of RANTES than control discs,20 making this an attractive candidate biomarker for examination in the current study. Catabolic activity leading to intervertebral disc tissue matrix turnover, including collagens and proteoglycans, is one of the hallmarks of intervertebral disc degeneration and may reflect active disease. Tissue inhibitor of metalloprotease (TIMP)-1, C-terminal telopeptide of type II collagen (CTX-II), and aggrecan chondroitin sulfate 846 epitope (CS846) were chosen as measures of the catabolic activity that affects disc matrix components. Matrix metalloproteases (MMPs) are a family of catabolic enzymes responsible for matrix protein breakdown, and endogenous inhibitors (TIMPS) control their activity.9,10,21 High serum MMP and TIMP levels are detectable in individuals with disease
NOVEMBER 2014–VOL. 62, NO. 11
JAGS
involving tissue turnover and remodeling, such as osteoarthritis and rheumatoid arthritis.22 TIMP-1 levels were therefore measured as a marker of anticatabolic activity. In addition to mediators involved in control of matrix homeostasis, the constituents of the matrix molecules themselves can be analyzed. Previous studies in articular cartilage have implicated CTX-II (a breakdown fragment of collagen II) and CS846 (a marker of aggrecan turnover) as important biomarkers detectable in serum of adults with osteoarthritis.23,24 In addition, high serum levels of CTXII, which increases with time after an annular injury in an animal model of intervertebral disc degeneration, have been demonstrated.25 Similarly, CTX-II is high in women with degenerative disc disease independent of knee or hand osteoarthritis26 and has been identified as a potential biomarker in disc degeneration.27 CTX-II was associated with disc space narrowing in a cross-sectional population-based study.28 CS846 has also been investigated in degenerative conditions and decreases with age in the lumbar intervertebral disc.29 Levels of CS846 are associated with risk of progression in rheumatoid arthritis.30 In addition to the effects on matrix, inflammatory and catabolic pathways are involved in the production of painproducing cytokines, representing a possible mechanistic link between these processes and the experience of pain. Pain biomarkers have been shown to be high in other painful degenerative conditions31 and are the third class of biomarkers evaluated in the current study. Neuropeptide Y (NPY) was chosen because of its important role in modulating interindividual variations in emotion and stress resiliency32 and the association between changes in NPY and changes in pain in subjects with neck and back pain.33 In addition, NPY has been identified in the annulus fibrosus in human intervertebral discs removed for back pain,34 underscoring its relevance in disc degeneration. Taken together, these candidate biomarkers provide assessment of inflammatory, catabolic, and pain states. It was hypothesized that the chosen biomarkers would be associated with pain and pain-related function in addition to any explanatory information that MRI provided. Imaging changes reflect a history of what has occurred to the intervertebral disc over time, and traditional imaging findings do not provide evidence of active matrix turnover, inflammation, or pain. To the contrary, serum biomarkers have the potential to provide evidence of active disease, which may have greater association with pain and pain-related function. Therefore, the objective of the current study was to examine the relationship between serum biomarkers and self-reported pain intensity and pain-related function in addition to the contribution of MRI findings of lumbar spine degenerative changes in older adults with chronic low back pain. Because imaging changes are ubiquitous in older adults with low back pain, it is hoped that examining the relationship between serum biomarkers and pain and pain-related function could lead to better metrics of active disease in older adults.
METHODS Recruitment English-speaking participants aged 60 and older with primarily axial low back pain, operationally defined as low
JAGS
NOVEMBER 2014–VOL. 62, NO. 11
back pain that is more severe than pain in other parts of the body, every day or almost every day for at least the previous 3 months were recruited from an academic medical center (n = 43). Individuals with a wide range of pain severity were recruited to facilitate examination of associations with biomarkers. Exclusion criteria were serious underlying illness, including fever, weight loss, recent change in character or intensity of pain, and neurological bowel or bladder dysfunction; acute low back pain flare; inability to undergo MRI because of claustrophobia, body mass index (BMI), metal objects, pacemaker, defibrillator, or neurostimulator; previous low back surgery; radiation of pain into the lower extremities beyond the gluteal fold, suggestive of radiculopathy; cognitive impairment (MiniMental State Examination score
OBJECTIVES: To examine the relationship between serum biomarkers and self-reported pain intensity and pain-related function, in addition to the contribution of magnetic resonance imaging (MRI) findings of lumbar spine degenerative changes, in older adults with chronic low back pain. DESIGN: Single-center cross-sectional cohort study. SETTING: Academic medical center. PARTICIPANTS: Individuals aged 60 and older with axial low back pain without radiculopathy or previously diagnosed osteoarthritis of the knee or hip or pain outside the low back that is more severe than the back pain (n = 43). MEASUREMENTS: To examine pain-related impairment, pain was measured on a pain thermometer and the McGill Pain Questionnaire Short Form was administered. To examine pain-related function or activity limitation, the Roland Morris Disability Questionnaire, Short Physical Performance Battery (SPPB), and repetitive trunk rotation were used. Single plasma samples were obtained before and after physical performance tests and analyzed for inflammatory markers (E-selectin and regulated on activation, normal T cell expressed and secreted (RANTES)), inhibitors of catabolic enzymes (tissue inhibitor of metalloproteinases-1 (TIMP-1)), markers of matrix turnover (C- telopeptide of type II collagen (CTX-II) and aggrecan chondroitin sulfate 846 (CS846)), and stress biomarkers (neuropeptide Y From the aDepartment of Physical Medicine and Rehabilitation; bFerguson Laboratory for Orthopaedic Research, Department of Orthopaedics; c Department of Medicine; dDepartment of Radiology; eDepartment of Psychiatry; fDepartment of Anesthesiology; gClinical and Translational Sciences Institute, University of Pittsburgh; and hGeriatric Research, Education and Clinical Center, Veterans Affairs Pittsburgh Healthcare System, Pittsburgh, Pennsylvania. Address correspondence to Gwendolyn Sowa, Associate Professor, Department of Physical Medicine and Rehabilitation, Co-Director, Ferguson Laboratory for Orthopaedic and Spine Research, University of Pittsburgh, 3471 5th Ave, Suite 201, Pittsburgh, PA 15213. E-mail: [email protected] DOI: 10.1111/jgs.13102
JAGS 62:2047–2055, 2014 © 2014, Copyright the Authors Journal compilation © 2014, The American Geriatrics Society
(NPY)). Conventional nongadolinium lumbar MRI was performed and analyzed quantitatively and clinically. RESULTS: Composite MRI measurements did not show significant correlation with pain or pain-related function. Basal levels and changes in serum biomarkers in response to activity, particularly NPY and RANTES, demonstrated associations with pain and pain-related function in addition to the explanatory power of MRI-based results. CONCLUSION: Serum biomarkers may be a metric for assessment of active disease in older adults, in whom imaging changes are ubiquitous. In addition, changing levels of biomarkers in response to activity suggests that they may be useful as metrics to measure treatment responses in future studies and may reflect potential targets for use in designing personalized treatment for older adults with low back pain. J Am Geriatr Soc 62:2047–2055, 2014.
Key words: low back pain; biomarkers; imaging
U
nderstanding the contribution of back pain to the function of older adults is of paramount concern because of the increasing prevalence of low back pain with age and expectations of older adults to remain active, but current tools to assess lumbar spine pathology focus on imaging findings and do not address pain or catabolic activity of the tissue. Although intervertebral disc degeneration is ubiquitous on magnetic resonance imaging (MRI) of older adults, only 42% report low back pain during the prior 12 months.1 The lack of correlation between anatomical changes and symptoms in older adults2,3 makes imaging a suboptimal predictive tool in identifying pain generators in older adults with low back pain. Despite the limitations of current imaging modalities, correct identification of pain generators is critical to guiding appropriate and successful treatments. This is of paramount concern in older adults, in whom multiple pain generators frequently coexist. Nevertheless, interventions
0002-8614/14/$15.00
2048
SOWA ET AL.
are frequently planned based on imaging studies, often resulting in unnecessary procedures and associated morbidity.4 Because serum levels of inflammatory and matrix breakdown products address disease activity in other musculoskeletal conditions,5,6 evaluation of biomarkers for low back pain has the potential to provide assessment of active pain generators in addition to the changes observable on imaging studies. Ascertaining which of these molecular pathways are linked to pain and pain-related functional limitations may differentiate individuals requiring therapeutic intervention from those in whom identification of alternative pain generators should be explored. In addition, evaluation of these markers has the potential to contribute to the understanding of the mechanism of the degenerative process and low back pain and may therefore lead to novel diagnostic and therapeutic tools. Candidate biomarkers chosen for this pilot study are important mediators in the inflammatory, catabolic, and pain pathways. Biological mediators of inflammation and catabolism are involved in intervertebral disc degeneration.7–10 In other degenerative conditions, inflammatory and catabolic biomarkers are sensitive indicators of disease activity and progression,5,6 and high levels of inflammatory cytokines and catabolic mediators have been isolated from degenerative discs.8,11,12 There are dramatic differences between the expression profiles of inflammatory and catabolic genes in an animal model of pathologic disc degeneration and a normal aging animal, with differences noted in critical inflammatory and catabolic genes.13 E-selectin and regulated on activation, normal T cell expressed and secreted (RANTES) were chosen as markers of inflammation. In addition to E-selectin’s known role as a marker of systemic inflammation, high levels are associated with positive straight leg raise in individuals with surgically confirmed nerve root compression14 and are a useful serum prognostic indicator in rheumatoid arthritis.15 More relevant to the current study, positive immunostaining for E-selectin has been found in surgically obtained herniated disc specimens, although all of these subjects had associated radicular pain.16 Nevertheless, the presence of positive staining in the disc suggests a possible role in axial low back pain as well. RANTES is expressed in response to inflammatory stimuli and affects catabolic activity by altering matrix metalloprotease production in articular cartilage.17 Human intervertebral discs also produce RANTES,18 serum levels have been shown to change in response to activity in other disease conditions,19 and painful discs express higher levels of RANTES than control discs,20 making this an attractive candidate biomarker for examination in the current study. Catabolic activity leading to intervertebral disc tissue matrix turnover, including collagens and proteoglycans, is one of the hallmarks of intervertebral disc degeneration and may reflect active disease. Tissue inhibitor of metalloprotease (TIMP)-1, C-terminal telopeptide of type II collagen (CTX-II), and aggrecan chondroitin sulfate 846 epitope (CS846) were chosen as measures of the catabolic activity that affects disc matrix components. Matrix metalloproteases (MMPs) are a family of catabolic enzymes responsible for matrix protein breakdown, and endogenous inhibitors (TIMPS) control their activity.9,10,21 High serum MMP and TIMP levels are detectable in individuals with disease
NOVEMBER 2014–VOL. 62, NO. 11
JAGS
involving tissue turnover and remodeling, such as osteoarthritis and rheumatoid arthritis.22 TIMP-1 levels were therefore measured as a marker of anticatabolic activity. In addition to mediators involved in control of matrix homeostasis, the constituents of the matrix molecules themselves can be analyzed. Previous studies in articular cartilage have implicated CTX-II (a breakdown fragment of collagen II) and CS846 (a marker of aggrecan turnover) as important biomarkers detectable in serum of adults with osteoarthritis.23,24 In addition, high serum levels of CTXII, which increases with time after an annular injury in an animal model of intervertebral disc degeneration, have been demonstrated.25 Similarly, CTX-II is high in women with degenerative disc disease independent of knee or hand osteoarthritis26 and has been identified as a potential biomarker in disc degeneration.27 CTX-II was associated with disc space narrowing in a cross-sectional population-based study.28 CS846 has also been investigated in degenerative conditions and decreases with age in the lumbar intervertebral disc.29 Levels of CS846 are associated with risk of progression in rheumatoid arthritis.30 In addition to the effects on matrix, inflammatory and catabolic pathways are involved in the production of painproducing cytokines, representing a possible mechanistic link between these processes and the experience of pain. Pain biomarkers have been shown to be high in other painful degenerative conditions31 and are the third class of biomarkers evaluated in the current study. Neuropeptide Y (NPY) was chosen because of its important role in modulating interindividual variations in emotion and stress resiliency32 and the association between changes in NPY and changes in pain in subjects with neck and back pain.33 In addition, NPY has been identified in the annulus fibrosus in human intervertebral discs removed for back pain,34 underscoring its relevance in disc degeneration. Taken together, these candidate biomarkers provide assessment of inflammatory, catabolic, and pain states. It was hypothesized that the chosen biomarkers would be associated with pain and pain-related function in addition to any explanatory information that MRI provided. Imaging changes reflect a history of what has occurred to the intervertebral disc over time, and traditional imaging findings do not provide evidence of active matrix turnover, inflammation, or pain. To the contrary, serum biomarkers have the potential to provide evidence of active disease, which may have greater association with pain and pain-related function. Therefore, the objective of the current study was to examine the relationship between serum biomarkers and self-reported pain intensity and pain-related function in addition to the contribution of MRI findings of lumbar spine degenerative changes in older adults with chronic low back pain. Because imaging changes are ubiquitous in older adults with low back pain, it is hoped that examining the relationship between serum biomarkers and pain and pain-related function could lead to better metrics of active disease in older adults.
METHODS Recruitment English-speaking participants aged 60 and older with primarily axial low back pain, operationally defined as low
JAGS
NOVEMBER 2014–VOL. 62, NO. 11
back pain that is more severe than pain in other parts of the body, every day or almost every day for at least the previous 3 months were recruited from an academic medical center (n = 43). Individuals with a wide range of pain severity were recruited to facilitate examination of associations with biomarkers. Exclusion criteria were serious underlying illness, including fever, weight loss, recent change in character or intensity of pain, and neurological bowel or bladder dysfunction; acute low back pain flare; inability to undergo MRI because of claustrophobia, body mass index (BMI), metal objects, pacemaker, defibrillator, or neurostimulator; previous low back surgery; radiation of pain into the lower extremities beyond the gluteal fold, suggestive of radiculopathy; cognitive impairment (MiniMental State Examination score
