Is neurogenic inflammation involved in tendinopathy? A systematic review

Neurogenic pain and inflammation have been hypothesised to play an important role in tendinopathy. This systematic review aimed to present and assess the evidence on neurogenic inflammation in tendinopathy. A systematic search was conducted through multiple databases to identify human case–control studies assessing neurogenic inflammation through the upregulation of relevant cells, receptors, markers and mediators. A newly devised tool was used for the methodological quality assessment of studies. Results were pooled based on the cell/receptor/marker/mediator assessed. A total of 31 case–control studies were eligible for inclusion. The tendinopathic tissue was obtained from Achilles (n=11), patellar (n=8), extensor carpi radialis brevis (n=4), rotator cuff (n=4), distal biceps (n=3) and gluteal (n=1) tendons. Through pooling the results of included studies based on the marker of neurogenic inflammation assessed, we identified possible upregulation of protein gene product 9.5 (PGP 9.5), N-methyl-D-aspartate Receptors, glutamate, glutamate receptors (mGLUT), neuropeptide Y (NPY) and adrenoreceptors in tendinopathic tissue versus control. Calcitonin gene-related peptide (CGRP) was not found to be upregulated, and the evidence was conflicting for several other markers. These findings show the involvement of the glutaminergic and sympathetic nervous systems and the upregulation of nerve ingrowth markers supporting the concept that neurogenic inflammation plays a role in tendinopathy.

No, if the assessor is aware of injury status or if no mention is made of assessor blinding.
N/A if only one group. 11 Method: Were the observational tests used to assess the main outcomes appropriate?
Yes, if the observational tests used were appropriate for the research question and the data with minimum 2 independent observers?
No, if no quantitative or semi-quantitative descriptive analysis was reported. 12 Method: Are the distributions of principal confounders in each group of subjects to be compared clearly described?
Yes, if summaries of participant age, sex, BMI (or height and weight) are provided for the case and control group.
No if study did not provide data for at least these factors. 13 Data Analysis: Are differences between neurogenic inflammation markers between the case and control group clearly reported?
Yes, if the comparison of neurogenic inflammation between case and control groups is clearly described. OR Yes, if data is provided in sufficient detail to calculate a comparison between the case and control groups No, if comparison is not clearly described.
No, if comparison is given as significant or nonsignificant without p-value or detailed data. N/A if only one group 14 Data Analysis: Does the study provide estimates of the random variability in the data for the main outcomes?
Yes, if an estimate of data variability is provided for sympathetic involvement. Acceptable estimates include SD, SE, and IQR.
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)   d) mGlur5 distinctive of late-stage tendinopathy, predominant on altered tenocytes and free nerve fibres in tendinopathy biopsies e) SP present on both peritendinous and tendon proper tissue f) SP on sprouting nerve fibres in 5 out 10 biopsies exhibiting signs of late stage tendinosis vs absent in control g) the occurrence and immunodensity of NMDAR1 correlated with that of SP in tendinopathic samples vs not so in control h) co-localisation between NMDAR1 and SP and phosphor-NMDAR1 and SP both in the tendinopathic and control biopsies, however only tendinopathic biopsies exhibited co-localisation of SP and phosphor-NMDAR1 within the tendon proper.

Bagge et al 2012
-ISH results--tendinosis tenocytes showed specific BDNF mRNA reaction -specific mRNA reactions were noted for tenocytes in non-tendinosis patients -IHC results--large number of tenocytes showed BDNF immunoreactivity in both tendinosis and non-tendinosis groups -BDNF is produced in the tenocytes of the human Achilles tendon, however BDNF immunolabelling and BDNF mRNA is not confined to all tenocytes in the Achilles tendon

Bjorklund et al 2011
-IHC-F results a) Difference in CB1 expression between groups was statistically significant (P<.05) with it being higher in the tendinosis group vs control

Xu et al 2011
-IHC results a) Immunoreactivity for PGP9.5 and GAP43 was rarely seen in the tendon tissue proper, but rather in the paratendinous tissue and endotenon between collagen bundles and near blood vessels BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) b) Large groups of nerve fascicles observed in torn and matched tendon groups vs control -GAP43 and PGP9.5 observed within tendon proper and or intimately associated with blood vessels c) Quantitative analysis showed that number of PGP9.5 and GAP43 immunoreactive nerves were significantly higher in matched subscapularis tendons vs control subscapularis tendons (P<0.05) and torn supraspinatus tendons (P<.0002 and P<.0001).

Schizas et al 2010
Quantitative Assessment 1) NMDAR1 a) No vessels within the tendon proper, b) vascular NMDAR1 in the tendon proper (5.9%) of tendinopathic group exhibited similar levels as in the paratendinous tissue c) 9 fold increase in vascular NMDAR 1 in tendinopathic tenocytes vs control 2) Glutamate a) Glutamate occurrence in vessels and cells was elevated 10 times in tendinopathic group b) Increased number of glutamate positive tenocytes in tendinopathic tendons vs control (p= 0.009) c) Vascular glutamate localised in the paratendinous tissue higher vs control 3) Correlation of NMDAR1 and glutamate occurrence: no correlation found in either 2 groups Combined IF and DAB staining 1) NMDAR1 a) Increased NMDAR1 immunoreaction in painful tendons, localisation of increased NMDAR1 immunostaining in tendinopathic samples b) Occurrence of NMDAR1 with PGP9.5-elevated in tendinopathy vs control 2) Glutamate a) Glutamate was elevated in tendinopathic tissue vs control b) Occurrence of glutamate with PGP9.5-in tendinopathy vs control c) Increased glutamate positive tenocytes in tendinopathic tissue vs control 3) Co-localisation of NMDAR1 and glutamate: a) Demonstrated in nerve fibres in all tendinopathic tissues vs none in control 14. Bagge et al 2009 There are marked immunoreactions for the neurotrophins NGF and BDNF and for the p75 receptor, but not for TrkA or TrkB, in the tenocytes of the human Achilles tendon 15. Bjur et al 2009 a) NPY-immunoreactions were seen in the nerve fascicles, and mildly in the perivascular nerve fibres, but none in the tenocytes. b) Y1 receptor-immunoreactions present in both non-tendinosis and tendinosis groups, seen in tenocytes and blood vessel walls -stronger immunoreactions present in tendinosis group vs control (p<0.01) c) Y2 Receptor-no immunoreactions in blood vessels wakks, tenocytes or nerve fascicles. -

Zeisiget al 2009
showed presence of catecholamine-synthesising enzyme TH in the fibroblasts of the tissue samples from 4/7 patients with TE and 2/4 patients with GE, and no detectable levels of this enzyme were found in fibroblasts of control tissue from the lateral epicondyle (0/6). -no evidence of such production in patients with TE or GE was found in the present study using staining for the ACh-synthesising enzyme ChAT. -no evidence of nerves positive for ChAT, whereas several nerve structures displaying TH-immunohistochemical reactions were detected. 17. Singaraju et al 2008 The IHC staining detecting CGRP and substance P was found globally throughout the tendon body in the proximal and distal sections of both groups with no significant differences between the control and experimental tendons.
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance -IHC results: General and Sensory Innervation Patterns 1) PGP 9.5-specific reactions for PGP 9.5 abundant in areas of loose CT 2) SP/ CGRP or SP-CGRP-LI were overall rarely detected in specimens of the tendinosis tendons; this corresponded to the normal tendon tissue proper 3) Sympathetic innervation patterns -Normal tendons: lower NPY and TH immunoreactive nerve fibres than in the loose paratendinous CT -Tendinosis Tendons: NPY & TH immunoreactive nerve fibres abundant in loose CT and around blood vessels -not a lot of difference btw normal and tendinosis samples 4) Adrenergic Receptors -nerve fascicles in the tendinous tissue displayed an immunoreaction pattern for adrenergic receptors similar to that of the loose paratendinous CT of controls, with greatest immunoreaction being for alpha 1 adrenoreceptor -alpha-1 adrenoreceptors in tendinosis tendon vasculature was more marked vs control by semiquantitative analysis -limited immunoreactions for alpha 2A adrenoreceptors in both healthy and tendinosis tissue -no general difference in the occurrence of beta1-adrenoreceptor in tendinosis vs control -tenocytes exhibited adrenergic receptor LI, alpha 2A adrenoreceptor-LI with more distinct immunoreaction compared to control 5) TH-like immunoreactions (-LI) in tenocytes -immunoreactions more distinct in tendinosis tendon samples vs normal 6) The amount of sympathetic innervation did not match the quantity of adrenergic receptors in the tendon tissue proper of the patellar tendon, particularly in tendinosis.
-These findings suggest that locally produced catecholamines can be mediators that bind to the frequently occurring adrenergic receptors

Results
o M2 receptor ▪ Immunoreaction in blood vessel walls observed in both groups, more pronounced in tendinosis group particularly in those with hyper cellularity ▪ No immunoreaction in tenocytes & nerve fascicles observed in controls ▪ Immunoreaction in tenocytes & nerve fascicles observed in specimens, particularly in those with hypercellularity o Choline acetyltransferase (ChAT) ▪ No immunoreaction in blood vessels or tenocytes in control ▪ Immunoreaction in blood vessels observed in tendinosis, particularly in those with profound hyper cellularity or abnormally appearing tenocytes ▪ No immunoreaction in tenocytes in specimens o Vesicular Acetylcholine transporter (VAChT) ▪ No immunoreaction in blood vessels or tenocytes in control ▪ Some immunoreaction in tenocytes seen in specimens, more so in those with hypercellularity or with abnormal appearance ▪ No immunoreaction in blood vessels in specimens o Acetylcholinesterase (AChE) Immunoreaction observed in blood vessels & tenocytes of both groups, with no convincing differences between both groups 24. Lian et al 2006 Semiquantitative Analysis tendon vs control -higher occurrence of SP= 0.567 -higher occurrence of PGP= 0.098 -lower occurrence of TH = 0.018

Bjur et al 2005
IHC results a) Innervation patterns-PGP9.5 was seen in tendinosis tissue, the staining was seen intimately associated with fine blood vessels unlike control b) Immunoreactions against CPRP and SP were also detected in thin nerve fascicles and as freely coursing nerve fibres, sometimes being closely located to fine blood vessels c) In normal tendon specimens, the immunoreaction for CGRP was more marked than that for SP d) CGRP/SP immunoreaction was only observed in the association with a subpopulation of the blood vessels BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)