Computed tomography osteoabsorptiometry-based investigation on subchondral bone plate alterations in sacroiliac joint dysfunction

Sacroiliac joint dysfunction (SIJD) is an underappreciated source of back pain. Mineralization patterns of the sacroiliac (SIJ) subchondral bone plate (SCB) may reflect long-term adaptations to the loading of the joint. Mineralization densitograms of 27 SIJD patients and 39 controls, were obtained using CT osteoabsorptiometry. Hounsfield unit (HU) values of the SCB mineralization of superior, anterior and inferior regions on the iliac and sacral auricular surfaces were derived and statistically compared between SIJD-affected and control cohorts. Healthy controls showed higher HU values in the iliac; 868 ± 211 (superior), 825 ± 121 (anterior), 509 ± 114 (inferior), than in the sacral side; 541 ± 136 (superior), 618 ± 159 (anterior), 447 ± 91 (inferior), of all regions (p < 0.01). This was similar in SIJD; ilium 908 ± 170 (superior), 799 ± 166 (anterior), 560 ± 135 (inferior), sacrum 518 ± 150 (superior), 667 ± 151 (anterior), 524 ± 94 (inferior). In SIJD, no significant HU differences were found when comparing inferior sacral and iliac regions. Furthermore, HU values in the inferior sacral region were significantly higher when compared to the same region of the healthy controls (524 ± 94 vs. 447 ± 91, p < 0.01). Region mineralization correlated negatively with age (p < 0.01). SIJD-affected joints reflect a high mineralization of the sacral inferior region, suggesting increased SIJD-related mechanical stresses. Age-related SCB demineralization is present in all individuals, regardless of dysfunction.

www.nature.com/scientificreports/ Previous studies showed complete SIJ non-conformity and denser mineralization patterns on the iliac side 12 .
High mineralization zones (≥ 700 HU) were found around the borders, apex and the corners of the auricular surfaces 12,20 . As asymmetrical gait and abnormal biomechanics are typically observed in SIJD patients, likely as a compensatory mechanism 21 , changes in bone mineralization patterns may reflect on the auricular surfaces caused by the abnormal loads instigated by SIJD. In unilaterally-affected SIJD, patients often load through their unaffected side for pain-relief leading to a decrease in muscle endurance and strength on the painful side which can affect the pelvic force-closure system 22,23 . This given study aimed to compare the mineralization distribution patterns of the dysfunctional SCB in the superior, anterior and inferior regions of both the sacrum and the ilium, by comparing that of a cohort of SIJD patients with that of an age-matched healthy control cohort with no history of LBP. The following hypotheses were investigated: 1. SIJD patients display higher mineralization patterns compared to the healthy state. 2. Non-affected contralateral SIJs in patients with unilateral SIJD display similar mineralization patterns to healthy controls. 3. Mineralization patterns in the SIJ are different between patients with unilateral and bilateral SIJD.

Materials and methods
Patients and controls. Twenty-seven patient cases diagnosed with unilateral or bilateral SIJD (13 females; 14 males; range 26 to 79 years) ( Fig. 1) were collected between 2009 and 2018 in the JCHO Sendai Hospital, Sendai, Japan. Two cases were used in a previous study on sacroiliac joint arthrodesis for SIJD relief 24 . All patients identified the posterior superior iliac spine as the main pain area by using their index finger (one-finger test) 25 and following thorough examination were considered to having SIJ pain. Definitive diagnosis of SIJD was confirmed by more than 70% pain relief at the SIJ region after SIJ local anesthetic injections under fluoroscopic guidance 26 . Patients with a history of infection, pathological conditions, tumors, cysts in the lumbopelvic area, recent lumbar spine and pelvic fractures, and seronegative spondylarthropathy were excluded. All included patients had a history of other injections including selective nerve root infiltration and/or lumbar disc nerve block that were negative. All patients were diagnosed as having severe chronic SIJD pain (for a minimum of 6 months) and reached the standard of indication for SIJ arthrodesis as ultima ratio: insufficient responsiveness to conservative treatments continued for longer than 6 months, difficulty in working, and/or marked restrictions of daily living due to recurrence of severe SIJ pain, even after undergoing repeated diagnostic/therapeutic injections and substantial physical therapy 24 . As a control cohort, 39 age-matched CT scans (20 females,19 males: range 21 to 82 years) acquired for clinical purposes to diagnose non-musculoskeletal pathologies, or to rule out injury related to acute trauma, were used. None of these cases had a current or past history of LBP, SIJ-related pathology or abnormalities on previous medical records. The Institute Review Board of JCHO Sendai Hospital, Sendai, Miyagi, Japan (no. 2019-1) approved the present study. Institutional approval was acquired for the use of patient datasets used in research studies for diagnostic and therapeutic purposes. Approval was granted on the grounds of existing datasets. Informed consent was obtained from all participants of this study. All methods were carried out in accordance with relevant guidelines and regulations.
Computed tomography osteoabsorptiometry. Data sets for CT-OAM were derived from conventional CT (SOMATOM as64 open, Siemens, Munich, Germany; Aquilion one, Toshiba, Tokyo, Japan). Slices thicknesses averaged 1.25 mm for the healthy patient CT and ranged from 0.7 to 5.0 mm thickness for the scans of the cases with SIJD. CT-OAM was evaluated using a specific image analysis program (Analyze, v7.4, Biomedical Imaging Resources, Mayo Foundation, Rochester, NY, USA). According to a previous study 12 , the sacral and iliac sides of each SIJ were manually segmented within the CT datasets before the data were false color-coded and superimposed on the 3-dimensionally reconstructed ilia and sacra for anatomical localization of the mineral  Analysis of densitogram patterns. The mineral density pattern of the iliac and sacral sides was evaluated based on the mean HU values of the regions on the densitogram for each dataset ( Fig. 2A). The auricular surfaces were subdivided into three regions: superior, anterior and inferior (Fig. 2B). These were defined as being three sections of equal size with the anterior region incorporating the apex of the auricular surface. The size of the tool was based on the size of the auricular surface of the specimen. Calculation of the mean HU value for each region was computed using non-calibrated CT grey values, obtained from a clinical-type scanner. These values were subsequently statistically compared between the different groups. For statistical analyses GraphPad Prism (version 8, San Diego, CA, USA) was used. Statistical significance was defined at the 5% (p ≤ 0.05) level. Outlier values were identified using Microsoft Excel (version 15.38, Redmond, WA, USA) and removed from the data. These were values (n) outside the range of the upper and lower bounds of 1.5 × the interquartile range (IQR): (n < 1st quartile − (1.5 × IQR); n > 3rd quartile + (1.5 × IQR)). Gaussian distribution was first assessed using a Shapiro-Wilk test. Depending on the distribution, a one-way ANOVA or a Kruskal-Wallis test with Dunn's post-hoc correction was undertaken for the multiple assessment of the data between the three regions. Mean HU values were reported ± standard deviation. Age correlations with mean HU values in the three regions between sexes, sides and within the bone were assessed using a two-tailed Spearman r test for non-parametric data or a two-tailed Pearson r test for parametric data. Correlation were defined as follows: strong ≥ 0.7, moderate 0.7 > r ≥ 0.5, weak 0.5 > r ≥ 0.3.
In the joints affected by unilateral SIJD and the contralateral unaffected sides, both the superior and the anterior regions of the ilium showed significantly higher HU values than the same regions of the sacrum (p < 0.05; Fig. 3B,D). In all SIJD cases, mean HU values of the inferior sacral region showed no significant differences when compared to the HU values of inferior region of the ilium (Fig. 4). In bilateral SIJD cases, HU values in both the anterior region and inferior region were high, and there were no significant differences when comparing the same regions of the ilium (p > 0.06; Fig. 3C). No significant differences in mean HU values were found between sexes (p > 0.6) nor sides (p > 0.1). See Supplementary Table S1 for all data. www.nature.com/scientificreports/ Differences in subchondral bone density exist in the inferior sacral region between SIJD cases and healthy controls. When comparing the mean HU values of the sacrum between all SIJD cases and healthy controls, SIJD showed significantly higher HU values in the inferior sacral region (p < 0.01). Regarding the three types of comparison as follows; healthy versus unaffected joint with contralateral SIJD, and the unilateral and bilateral SIJD cases, there were no significant differences in any regions on either bone (Fig. 5).
When comparing the healthy and unilateral SIJD joints, mean HU values of the inferior region of the sacrum in  www.nature.com/scientificreports/ unilateral SIJD was higher than healthy controls but, it was not statistically significant. Only when comparing the inferior region of the sacrum between healthy and bilateral SIJD were the mean HU values significantly different (p < 0.01; Fig. 5).

Discussion
This given study quantified subchondral bone mineralization in the sacroiliac joint using CT-OAM in two patient cohorts: healthy SIJ controls and SIJD sufferers. It provides first insights into morpho-mechanic differences related to SIJD and densitograms of the sacral and iliac SCB based on mean HU values in the superior, anterior and inferior regions of the SIJ auricular surfaces. The patterns presented here are likely the result of chronically recurring loading conditions of individuals with and without SIJD, which represent the biomechanical stresses applied to the SIJ. In healthy SIJs, the sacral inferior region showed lower mineralization values than that of the ilium. In SIJD, values of the sacral inferior region tended to be higher and closer to those of the ilium. Particularly in bilateral SIJD, in addition to the sacral inferior region, the anterior region HU values were higher. These are the most specific mineralization patterns of SIJD-affected joints. Thus, our first hypothesis; dysfunctional SIJs display higher mineralization patterns compared to the healthy state, can be confirmed in the inferior sacral region. The dysfunctional joint might be stressed in the sacral inferior region or, vice versa, stresses exerted to the inferior portion may be causative for SIJD. Anatomically, the posterior SIJ forms a syndesmosis. The anterior auricular www.nature.com/scientificreports/ region, however, consists of synovial tissues and a joint capsule like other typical synovial joints 27 . In an early stage of sacroiliitis, inflammation is often detected in the inferior portion of the joint 28 . Because the inferior portion of the joint is rich in synovial tissues, this indicates that the inferior region could be more mobile than the superior region. A recent study revealed that the upper portion of the SIJ was stressed in stance phase and the inferior portion was stressed in swing phase of bipedal walking 29 . The superior portion may work primarily as a shock-absorber and the inferior portion as a 'slider' for smooth bipedal walking. In cases of dysfunction, abnormal sliding of the joint surface may cause increased stresses in the inferior SIJ region. The sacral cartilage is thick originally and would contribute to stress-reduction of the sacral SCB in stance phase, but, in the inferior sacral region, the cartilage is thinner 30 . Therefore, in SIJD, the alterations in density of the inferior SCB could result from abnormal moving of the joint. Furthermore, SIJ ligaments contribute to pelvic stability [31][32][33] and can cause LBP 34,35 . In non-painful pelvic biomechanics, the pelvic musculature and surrounding soft tissues optimize the load-transmission through the cortical shell 36 . Increased mineralization in the inferior region may reflect impaired force-transmissions via the pelvic musculature and ligaments, perhaps laxity-related 37 . This may be the influence of SIJD, or it may reflect a ligamentous dysfunction elsewhere around the kinematic chain involving the inferior portions of the interosseous, posterior and the long SIJ ligaments, as well as the proximal parts of the sacrotuberous and sacrospinous ligaments 27,31 . Furthermore, sacral asymmetry had a higher prevalence in SIJD cases which might contribute to asymmetrical loading throughout life and might subsequently affect overall mineralization 38 .
No differences were found in any region on either bone when comparing the controls with the unaffected side of patients with unilateral SIJD, suggesting that SCB-mineralization is unaltered by preferential loading to the unaffected side 22,23 . When subjected to bilateral pain, it seems impossible to relieve the pain through compensatory mechanisms of a contralateral unaffected side, therefore, it is possible that both sides succumb to an increase in loads, specifically in the sacral inferior region, and could cause the increase in mineralization in this region. Thus, our second hypothesis; non-affected SIJs in patients with SIJD display similar mineralization patterns to healthy controls, can be approved. These findings could also suggest inflammation of tissues within the SIJ. This hypothesis has previously been rejected as inflammation and subsequent osteoblastic activity within the joint, would be observed on CT or bone-scans similar to sacroiliitis and ankylosing spondylitis, which has not been reported 8,9 . However, Maigne et al. 8 reported an increased bone uptake of the subchondral lamella in the painful side of patients with SIJD, when using bone-scans. Although their results had poor sensitivity, it was suggested that this bony increase could reflect changes without an acute inflammatory episode 8 . In this case, the mineralization increases of the auricular surface may reflect low-grade inflammation of tissues which locally, may cause bone formation as seen in cases of aseptic loosening in hip arthroplasty 39 . www.nature.com/scientificreports/ No significant change in patterns were found between patients with unilateral or bilateral SIJ pain, thus our third hypothesis; different mineralization patterns in the SIJ would be observed when comparing unilateral SIJD to bilateral SIJD joints, is rejected. Mineralization-induced density remained unaffected by altered biomechanics caused by unilateral pain, but it is impossible to know how many of the patients realistically exhibited altered loading in response to SIJD. Mean SCB density values were neither sex-nor side-dependent. Therefore, SCB mineralization is not related to side-nor sex influences in the loading of the posterior pelvis, occurring as a consequence of a different osteoligamentous alignment 31,33,40 . In all cohorts, negative age correlations were found when comparing the density values of each region, especially in the superior and inferior regions reflecting the osteopenia the bone succumbs to with time, caused by age-related changes in osteoanabolic metabolism 41 .
Regarding the limitations of the study, healthy scans were different patients to the SIJD cohorts; therefore, this comparison did not take into account population differences and potential variables between the two cohorts except, age, sex and SIJ pathologies. Size differences in the regions may have also been an influencing factor of the results as the regions were based on the size and shape of the auricular surface which would have been different for each individual. Furthermore, regional analysis may also have included some pixels outside of the auricular surfaces (HU value of 0), which may have caused an underestimation of the regional HU mean. In addition, the manual isolation of the three regions was performed by only one author. Therefore, the reproducibility of outlining the exact same regions was not tested. However, the regions were defined as being three sections of equal size with the anterior region incorporating the apex of the auricular surface. These are easily reproducible on each specimen regardless of the shape of the auricular surface.

Conclusions
Joints with SIJD reflect increased stresses in the inferior portion of the SIJ, resulting from mechanically induced increases in density. Interregional comparisons reflect mineralization differences between healthy controls and SIJD. Subchondral bone mineralization is independent of sexes and sides but shows age-related morphomechanical alterations.

Data availability
The data acquired in the course of this study are available from the corresponding author on request.