To determine differences in the metabolism of proteoglycans and the gene expression of proteinases and their inhibitors between patellar tendons exhibiting chronic overuse tendinopathy and normal patellar tendons in humans.
To determine differences in the metabolism of proteoglycans and the gene expression of proteinases and their inhibitors between patellar tendons exhibiting chronic overuse tendinopathy and normal patellar tendons in humans.
Rates of loss and synthesis of proteoglycans were determined. Radiolabeled and total proteoglycans retained in and lost from the tissue were analyzed by fluorography and Western blotting. Levels of messenger RNA for matrix metalloproteinase 1 (MMP-1), MMP-2, MMP-3, MMP-9, MMP-13, ADAMTS-1, ADAMTS-4, ADAMTS-5, tissue inhibitor of metalloproteinases 1 (TIMP-1), TIMP-2, TIMP-3, and TIMP-4 were determined in fresh tissue.
The rate of loss of 35S-labeled proteoglycans was greater in abnormal tendons, as was the rate of synthesis of proteoglycans. Fluorography and Western blotting revealed the presence of greater amounts of large proteoglycans (aggrecan and versican) in abnormal tendons, and these proteoglycans were rapidly lost from the matrix of abnormal tendons. There was no significant difference in the expression of ADAMTS-1, ADAMTS-4, ADAMTS-5, MMP-1, MMP-2, MMP-3, MMP-13, TIMP-2, TIMP-3, or TIMP-4. There was a significant increase in the expression of MMP-9 and TIMP-1 in abnormal tendons.
Our findings suggest that a change in the proteoglycan content of the extracellular matrix in abnormal tendons results from the altered metabolism of the cells, reflected in the enhanced synthesis of the large proteoglycans aggrecan and versican, and does not appear to result from changes at the level of gene expression.
Overuse patellar tendinopathy can contribute to 30% of injuries in individuals participating in exercise and sports (1). Patellar tendinopathy is characterized by proximal patellar tendon pain and tenderness that inhibits an individual's ability to exercise. The predominant feature of tendinopathy is a change in the appearance and organization of the extracellular matrix of the tissue, especially the disorganization of collagen fibers (2, 3).
Matrix disorganization may be the result of degenerative changes or a failed healing process associated with mechanical overloading (4). Degenerative changes may be at the end of a continuum of change that is proposed to include early cell-driven changes (reactive tendinopathy) through increasing extracellular matrix disorganization (tendon disrepair and degenerative tendinopathy) (5). Although the condition is likely to be driven by local levels of cytokines, tendinopathy does not appear to involve wider inflammatory processes (4, 6).
In normal tendons, the tenocytes are embedded in an extracellular matrix made up of type I collagen fibers arranged in parallel bundles with smaller amounts of other collagens, proteoglycans, hyaluronan, and noncollagenous proteins. Proteoglycans play a role in tissue hydration and regulate collagen integrity, while collagen provides the tissue with its tensile strength. Tenocytes are responsible for the synthesis and degradation of all of the macromolecular components of tendon. In normal tendons, the metabolism of tenocytes is reflective of the maintenance of the extracellular matrix where there is a balance between synthesis and catabolism of matrix macromolecules. Yet normal tendon has a limited potential to respond to changes in mechanical loading and to initiate repair of the extracellular matrix in response to overloading.
We have previously demonstrated increased levels of the large aggregating proteoglycans, aggrecan and versican, and of the small proteoglycans, biglycan and fibromodulin, in the extracellular matrix in human patellar tendinopathy compared with normal tissue (7). Furthermore, both the aggrecan and versican macromolecules in the matrix of abnormal tissue are present in significantly degraded states (7). In that study (7), no significant change was observed in the levels of type I collagen within the extracellular matrix of the abnormal tissue, nor were there significant changes in the expression of genes for the major collagens and proteoglycans associated with the tissue.
Previous studies of normal and abnormal human Achilles tendons have investigated gene expression for major matrix macromolecules and proteinases responsible for the catabolism of matrix molecules (8–10). This led to the suggestion that tendinopathy may result from a change in the metabolism of the tissue (11, 12).
No previous studies have directly investigated the dynamics of the metabolism of matrix macromolecules in normal and diseased human tendons. In this study, we determined the rates of catabolism and synthesis of newly synthesized proteoglycans in normal human patellar tendons and those with chronic overuse tendinopathy. The present study also established expression of matrix proteinases (matrix metalloproteinase [MMP] and ADAMTS) as well as tissue inhibitor of metalloproteinases (TIMP) in normal and abnormal tendons.
Normal human tendon tissue was obtained from patients undergoing anterior cruciate ligament reconstruction using a patellar tendon graft. Abnormal tendon tissue was obtained from patients undergoing surgical tendon debridement for recalcitrant overuse patellar tendinopathy. A section of tendon that was a minimum of 15 mm in length was harvested from the patellar enthesis in both groups. All patients gave written, informed consent, and all research was performed with the approval of the La Trobe University Human Ethics Committee.
Normal and abnormal tendons were surgically removed from patients, cut into small pieces (1–2 mm2), and washed in sulfate-free medium as previously described (13). Samples (100–200 mg wet weight of tissue) were then placed in sulfate-free medium containing 150 μCi/ml of 35S-sulfate at 37°C for 6 hours (13). Tendon samples were then washed with Dulbecco's modified Eagle's medium (DMEM) containing 1 mg glucose/ml buffered with 15 mM HEPES, 10 mM BES, and 10 mM TES (pH 7.2) to remove any of the unincorporated 35S-sulfate. DMEM contains sufficient chemical levels of sulfate to aid in the removal of unincorporated 35S-sulfate (14). Duplicate samples containing 50–100 mg wet weight of tendon were placed into sterile plastic vials that contained 4 ml of medium, for periods of up to 10 days. Medium was collected every 24 hours and replaced with fresh medium. The spent medium was stored at −20°C in the presence of proteinase inhibitors (15). The proteoglycans were extracted with 4 ml of 4M guanidine hydrochloride (GuHCl) containing 0.1M disodium sulfate and 0.1% (volume/volume) Triton X-100 and buffered with 0.05M sodium acetate (pH 6.1) (GuHCl buffer) in the presence of proteinase inhibitors at 4°C for 72 hours, followed by 4 ml of 0.5M NaOH at 21°C for 24 hours (13).
The amounts of radiolabeled proteoglycans present in the medium on each day of culture and in tissue extracts at the end of the culture period were determined using PD-10 columns (13). The rate of loss of 35S-labeled proteoglycans from the tendon explants was calculated from the amount of 35S-labeled proteoglycans that was present in the medium on each day of incubation and that remaining in the matrix at the end of the culture period (14). These data enabled the percentage of 35S-labeled proteoglycans remaining in the matrix on each day to be determined, and these results were plotted as a function of time in culture (13). This analysis was repeated using tissue from 5 normal subjects and 5 subjects with tendinopathy.
Normal and abnormal tissue samples were incubated with 35S-sulfate for 6 hours, as described above. A single batch of medium containing 35S-sulfate was used for all incubations. At the end of the incubation period, proteoglycans were extracted with GuHCl and NaOH, and the amounts of radiolabeled proteoglycans present in the medium and tissue extracts at the end of the culture period were determined, as described above. From these data, the amount of 35S-sulfate incorporated into proteoglycans was calculated and expressed as disintegrations per minute per 100 mg wet weight of tissue per 6 hours (16).
Following incubation with 35S-sulfate, tendon explants were cultured for periods of up to 10 days as described above. Proteoglycans present in the media and GuHCl extracts were purified using ion-exchange chromatography (Q-Sepharose columns; Pharmacia) (17). The proteoglycans were then deglycosylated with chondroitinase ABC and keratanase, as previously described (17), prior to electrophoresis on 4–15% gradient polyacrylamide–sodium dodecyl sulfate gels, and the gels were either subjected to fluorography or electroeluted onto polyvinylidene difluoride membranes (Immobilon-P membranes; Millipore) followed by Western blotting as previously described (7). For fluorography, the gel was placed in Amplify (Amersham Biosciences) for a period of 30 minutes and dried prior to exposure to x-ray film at −80°C for up to 60 days (18).
For Western blotting, the following antibodies were used: LF-99 (N-terminus of human versican), LF-136 (N-terminus of human decorin) (both kindly supplied by Dr. L. Fisher, Bone Research Branch, National Institute of Dental Research, Bethesda, MD), JD5 (G3 domain of human aggrecan; kindly supplied by Dr. J. Dudhia, Royal Veterinary College, Hatfield, UK), and N-14 (N-terminus of human fibromodulin; Santa Cruz Biotechnology). Antibodies were detected using an enhanced chemiluminescence detection kit, and secondary antibodies (raised against rabbit and mouse antibodies) were from Chemicon. This was repeated using tissue from 5 normal subjects and 5 subjects with overuse tendinopathy.
Samples (50–150 mg) of normal tendon (n = 9) and abnormal tendon (n = 12) were separately homogenized in 1.5 ml PureZOL/100 mg tissue (Bio-Rad), and total RNA was isolated using a kit according to the recommendations of the manufacturer. The procedure included digestion of genomic DNA with DNase I. The mean ± SD absorbance ratio A260/A280 was 1.6 ± 0.05 for normal tendons and 1.7 ± 0.1 for abnormal tendons. Normal tendons yielded 20–70 ng RNA/mg wet weight of tissue, and abnormal tendons yielded 85–330 ng RNA/mg wet weight of tissue. These results are consistent with the low cellularity of normal tendons compared with abnormal tendons (8, 9, 19).
Complementary DNA (cDNA) was synthesized using a reverse transcription kit from Bio-Rad. The resulting cDNA (∼9 ng/sample) was subjected to real-time PCR amplification in an iCycler iQ Detection System. The primers (see below) were designed using Beacon Designer 2.
The real-time PCR conditions were described previously (7). The values obtained for messenger RNA (mRNA) for a gene were normalized to GAPDH mRNA using the formula: gene/GAPDH = 2[Ct(GAPDH) − Ct(gene)] (20). In order to account for potential changes in GAPDH mRNA levels (8), the samples were also normalized to 18S ribosomal RNA (18S rRNA), and a similar outcome was obtained (data not shown). The primer pairs were as follows: for human MMP-1 (accession no. NM_002421; 135 bp), forward 5′-CTG-CTG-CTG-CTG-CTG-TTC and reverse 5′-AAC-TTG-CCT-CCC-ATC-ATT-CTT; for human MMP-2 (accession no. NM_004530; 124 bp), forward 5′-TGA-CGG-TAA-GGA-CGG-ACT-C and reverse 5′-ATA-CTT-CAC-ACG-GAC-CAC-TTG; for human MMP-3 (accession no. NM_002422; 101 bp), forward 5′-ACA-AGG-AGG-CAG-GCA-AGA-CAG and reverse 5′-GCC-ACG-CAC-AGC-AAC-AGT-AG; for human MMP-9 (accession no. NM_004994; 143 bp), forward 5′-TGA-CAG-CGA-CAA-GAA-GTG and reverse 5′-CAG-TGA-AGC-GGT-ACA-TAG-G; for human MMP-13 (accession no. NM_002427; 107 bp), forward 5′-CCA-GGC-ATC-ACC-ATT-CAA-G and reverse 5′-CAT-CTT-CAT-CAC-CAC-CAC-TG; for human ADAMTS-1 (accession no. NM_006988; 124 bp), forward 5′-CGA-AGA-CGA-GGA-CGA-AGG and reverse 5′-ACA-TAG-CGG-TGA-CTG-GAC; for human ADAMTS-4 (accession no. NM_005099; 142 bp), forward 5′-ACC-GAA-GAG-CAC-AGA-TTC and reverse 5′-ATG-AGG-CAG-CAA-CAG-AAG; for human ADAMTS-5 (accession no. NM_007038; 119 bp), forward 5′-ACC-GAT-GGC-ACT-GAA-TGT-AG and reverse 5′-ACT-CCG-CAC-TTG-TCA-TAC-TG; for human TIMP-1 (accession no. NM_003254; 141 bp), forward 5′-TGT-TGC-TGT-GGC-TGA-TAG and reverse 5′-GCT-GGT-ATA-AGG-TGG-TCT-G; for human TIMP-2 (accession no. NM_003255; 109 bp), forward 5′-GCA-GGA-GGA-ATC-GGT-GAG and reverse 5′-AGA-ACA-GGC-AAG-AAG-CAA-TG; for human TIMP-3 (accession no. NM_000362; 148 bp), forward 5′-GCG-TCT-ATG-ATG-GCA-AGA-TG and reverse 5′-AAG-CAA-GGC-AGG-TAG-TAG-C; for human TIMP-4 (accession no. NM_003256; 145 bp), forward 5′-TTA-CCA-GGC-TCA-GCA-TTA-TGT-C and reverse 5′-GGA-CTC-TTG-AAG-GGA-TGT-GAT-G; for human GAPDH (accession no. DQ403057; 138 bp), forward 5′-GGC-ACC-GTC-AAG-GCT-GAG-AAC and reverse 5′-GGT-GAA-GAC-GCC-AGT-GGA-CTC; and for human 18S rRNA (accession no. NG_002801; 150 bp), forward 5′-AAC-GAT-GCC-AAC-TGG-TGA-TGC and reverse 5′-CTC-CTG-GTG-GTG-CCC-TTC-C.
The Mann-Whitney test was used to determine the level of significance. P values less than 0.05 were considered significant. All analyses were performed using SPSS, version 14.0.
After 10 days in culture, ∼90% of the 35S-labeled proteoglycans remained in the matrix of normal tendons compared with 40% in the matrix of abnormal tendons (Figure 1). This indicates a faster rate of loss of radiolabeled proteoglycans from the matrix of abnormal tissue. This analysis was repeated using tissue from 5 normal subjects and 5 subjects with patellar tendinopathy. Similar rates of loss of radiolabeled proteoglycans, with a variation of <5%, were observed.
When determining the rate of loss of radiolabeled sulfate, it became evident that the incorporation of radiolabeled sulfate into proteoglycans was 20–25 times greater in all abnormal tendon samples than in normal ones. These data were obtained from tissue samples incubated with different batches of medium containing radiolabeled sulfate. In order to confirm this, the rate of incorporation of radiolabeled sulfate into proteoglycans in normal and abnormal tendons was determined under controlled conditions in which the same batch of medium containing 35S-sulfate was used. The rate of incorporation of 35S-sulfate into proteoglycans was 25-fold greater in abnormal tendons (1.432 × 106 dpm/100 mg wet weight per 6 hours) than in normal tendons (0.056 × 106 dpm/100 mg wet weight per 6 hours).
Normal tendons had low levels of 35S-labeled proteoglycan core proteins in the matrix of tissue immediately after incubation with 35S-sulfate (Figure 2, lane d). The majority of the radiolabel was associated with proteoglycan core proteins of 40–43 kd. This is consistent with the synthesis of the small leucine-rich proteoglycans decorin and biglycan by the tissue (13). The levels of radiolabeled proteoglycans remaining in the matrix decreased with time in culture. Considerably lower levels of radiolabeled molecules were observed in the medium compared with the levels present in the matrix of the cultures (Figure 2, lanes a–c, e, and f).
In contrast, in abnormal tendons the majority of the radiolabel present immediately after incubation with 35S-sulfate was associated with large proteoglycan core proteins of >200 kd. This indicates that the enhanced rate of proteoglycan synthesis in abnormal tissue described above was associated with the increased synthesis of the large proteoglycans aggrecan and versican (Figure 2, lane j). Small amounts of radiolabel were associated with bands corresponding to the core proteins of small proteoglycans (∼40 kd). The levels of these large and small radiolabeled core proteins that were present in the matrix rapidly decreased with time in culture. Most of the radiolabeled macromolecules present in the medium of cultures of abnormal tissue were associated with the large proteoglycan core proteins and their fragments (∼80–>200 kd), and these decreased in intensity over the culture period (Figure 2, lanes g–i). No radiolabeled bands corresponding to the core proteins of small proteoglycans were evident in the medium of cultured abnormal tendons (Figure 2, lanes g–i). These results were confirmed using tissue from 5 normal subjects and 5 subjects with overuse tendinopathy. A similar distribution of radiolabeled core proteins was observed.
Figure 3A shows a Western blot of proteoglycan core proteins isolated from a normal human patellar tendon probed with an antibody to the N-terminus of the core protein of versican. In the matrix of normal tendon on day 0 (Figure 3A, lane d), a prominent versican core protein fragment of ∼50 kd was present. No versican core proteins were evident in the medium of cultures of normal tendon. Higher levels of versican fragments, however, were present in the matrix and medium of abnormal tissue (Figure 3A), and this is consistent with previous observations of increased levels of versican in abnormal human tendons (7, 21). In the medium of abnormal tissue, a band of >200 kd was present, and this likely represents the intact core protein of versican (Figure 3A, lane g). The intensity of this band decreased with time in culture, as did the intensity of another prominent versican core protein fragment of ∼60 kd. This versican fragment was also present in the matrix of abnormal tendon cultures.
Analysis with an antibody to the C-terminus of aggrecan did not show significant aggrecan fragments in the medium of normal tendon (Figure 3B, lanes a–c). The matrix of normal tendon contained only low molecular weight aggrecan fragments with a prominent fragment of ∼55 kd, the intensity of which did not change with time in culture (Figure 3B, lanes d–f). It was evident that the matrix and medium of abnormal tissue contained more aggrecan than did normal tendon. In the matrix of abnormal tissue, the band of >200 kd is likely to be the intact core protein of aggrecan, while the bands corresponding to aggrecan fragments migrated at ∼80 kd, ∼90 kd, and ∼100 kd (Figure 3B, lanes j–l). With time in culture these bands in the matrix and medium decreased in intensity (Figure 3B, lanes g–i).
Figure 3C shows immunodetection with an antibody to the N-terminus of decorin. In both normal and abnormal tissue, core proteins corresponding to decorin (∼40–43 kd) were present in the matrix, and there was little change in the intensity of these bands over the culture period (Figure 3C, lanes d–f and j–l). Neither the culture medium of normal nor abnormal tissue contained decorin core proteins (∼40 kd) (Figure 3C, lanes a–c and g–i).
When an antibody to the N-terminus of fibromodulin was used, a single band of ∼60 kd was observed in the matrix and medium in both normal and abnormal tendons (Figure 3D). The intensity of this band did not change with time in culture. The only difference was that a greater amount of this proteoglycan was associated with the abnormal tissue. Similar results were obtained when the experiment was repeated using samples from 5 normal subjects and 5 subjects with overuse tendinopathy.
Expression of mRNA for MMP-9 (Figure 4A) and TIMP-1 (Figure 4C) was significantly up-regulated in abnormal tendons compared with normal tendons. A trend toward a down-regulation of expression of MMP-3 was seen in abnormal tendons (P = 0.06) (Figure 4A). There were no significant changes in ADAMTS-1, ADAMTS-4, or ADAMTS-5 gene expression (Figure 4B).
The rate of synthesis of proteoglycans was ∼25-fold greater in abnormal patellar tendons than in normal tissue. This was due to the significant increase in the synthesis of the large proteoglycans aggrecan and versican in abnormal tissue, whereas negligible levels of these proteoglycans were synthesized in normal tissue (Figure 2). Normal tissue predominantly synthesized small leucine-rich proteoglycans (Figure 2). The increase in the synthesis of large proteoglycans by abnormal tissue was reflected in greater tissue levels of large proteoglycans in the matrix of tendons in human patellar tendinopathy (Figure 3). This is consistent with the results of previous studies showing increased levels of large proteoglycans associated with overuse tendinopathy (7, 21). In addition, the higher cellularity of the abnormal tissue is likely to contribute to the higher incorporation of the radiolabel observed in this tissue (discussed below).
The overall rate of loss of 35S-labeled proteoglycans from the extracellular matrix of abnormal human patellar tendons was greater than the rate measured in normal tendons (Figure 1). This was primarily due to the loss of the radiolabeled large aggregating proteoglycans, which were more abundant in the abnormal tissue, and the majority of these were lost in the first 4 days of the culture period. These results do not differ from those in normal tissue where large proteoglycans are catabolized rapidly in normal tendons and other dense fibrous connective tissues (13, 22). In contrast, small proteoglycans are lost at a much slower rate in these tissues, which was also observed in both abnormal and normal tendons in the present study. In abnormal samples, the rate of loss of small proteoglycans was detected after the first 4 days of culture (Figures 1 and 2). These results indicate that the main difference between the normal and abnormal tendons is in the rate of synthesis of large proteoglycans in particular, whereas the rates of loss of both the large and the small proteoglycans follows the pattern observed in the normal tissue.
This study also shows that large proteoglycans are rapidly degraded so that mainly fragments remain in cultured tissue (Figures 2 and 3). To elucidate whether in overuse patellar tendinopathy there is a change in the gene expression of matrix-degrading enzymes and their inhibitors, which have been shown to or are likely to play a part in the proteolytic processing of proteoglycans (10, 17, 23, 24), the expression of genes for a number of proteinases (MMPs and ADAMTS) and their tissue inhibitors (TIMPs) was determined in normal and abnormal tissue. Of these, only MMP-9 and TIMP-1 were found to have been significantly up-regulated in abnormal tendons (Figure 4). This suggests that the proteolysis of proteoglycans in the abnormal tissue is not likely to have been affected at the gene expression level of proteinases and their inhibitors.
These observations are consistent with those of Jones et al (9), who showed that there was no difference in the expression of ADAMTS-1, ADAMTS-4, and ADAMTS-5 genes between human Achilles tendons with chronic pain and normal tendons. Corps et al (10) also reported that there was no significant change in the expression of ADAMTS-4 between normal Achilles tendons and Achilles tendons with chronic pain. In the latter study, Corps et al analyzed the nature of ADAMTS-4 that was present in the matrix and showed an increase in the level of the processed active form of this proteinase in painful Achilles tendons. Taken together, the results of the present study and those of the studies by Jones et al (9) and Corps et al (10) support the notion that in painful tendons ADAMTS-4 activity is elevated by the activation of the proteinase and not by increased gene expression.
It should be noted that Jones et al (9) and Corps et al (10) showed that in ruptured Achilles tendons there was a significant up-regulation of the expression of ADAMTS-4 compared with that in painful or normal Achilles tendons, and that Western blot analysis showed that the mature, inactive form of ADAMTS-4 was present in the matrix (10). In the present study, the elevated expression of MMP-9 and TIMP-1 genes in abnormal tendons likely reflects changes in collagen metabolism in abnormal tendons. Jones et al (9) showed no changes in the expression of MMP-9 and TIMP-1 in painful Achilles tendons, but a significant increase in the expression of these genes, in addition to a significant increase in MMP-1 and MMP-19 expression, was observed in ruptured Achilles tendons.
A number of studies have shown that increased cellularity is associated with overuse tendinopathy (7, 10, 19, 20), and this would explain the increase in total levels of mRNA measured in abnormal tendons in the present study. It is likely that, as the result of the underlying pathologic process, more cells are present in the tissue, thereby producing a net increase in tissue proteoglycans. However, the results of the analysis of the radiolabeled proteoglycans that were synthesized by abnormal tissue strongly suggest that more large proteoglycans are synthesized and retained within the matrix of abnormal tissue than normal tissue. Since gene expression of large proteoglycans does not differ significantly between normal and abnormal tendons (7), this suggests that transcriptional regulation may not play a direct role in the pathologic mechanism, and that regulation may be occurring at the level of translation of the appropriate mRNA into protein. Regulation of catabolism appears to be the outcome of the activation of ADAMTS and possibly MMPs. The details of the mechanism of these regulatory processes warrant further investigation.
The absence of inflammatory cells has brought into question the role of inflammation in the pathologic mechanism, but it is likely that changes in the phenotype and metabolism of tendon cells in overuse tendinopathy may be driven by locally produced cytokines. Indeed, increased synthesis of proteoglycans in abnormal tendons may be driven by increases in growth factors and cytokines such as interleukin-1α (IL-1α), IL-1β, transforming growth factor β, and/or tumor necrosis factor α (25–27), possibly as a result of repetitive mechanical stress. These growth factors and cytokines may also stimulate the activation of matrix-degrading enzymes such as ADAMTS and MMP, eventually leading to the breakdown and disorganization of the extracellular matrix that is observed in overuse tendinopathy (28). Notably, in the present study, tissue levels of the small proteoglycans decorin and fibromodulin were maintained throughout the culture period in abnormal tendons (Figures 3C and D, lanes d–f and j–l), whereas tissue levels of the large proteoglycans decreased with time in culture (Figures 3A and B, lanes d–f, and j–l).
Overall, these data support the concept that tendons exhibiting overuse tendinopathy are characterized by an altered extracellular matrix that has a different composition, structure, and metabolism compared with normal tendons, and that these changes result in part from the increased synthesis of large proteoglycans that are rapidly lost from the extracellular matrix of abnormal tendons. These findings indicate that the extracellular matrix of the tendon exhibiting overuse tendinopathy is a dynamic structure characteristic of an attempted adaptive or healing process.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Samiric had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Parkinson, Samiric, Ilic, Cook, Feller, Handley.
Acquisition of data. Parkinson, Samiric, Ilic, Cook, Feller, Handley.
Analysis and interpretation of data. Parkinson, Samiric, Ilic, Cook, Feller, Handley.