ABSTRACT
Purpose:
Diagnosis of partial rotator cuff tears and tendonopathy using conventional MRI has provenvariable. Quantitative T2 mapping may have application for assessing rotator cuff health. In order toevaluate the usefulness of T2 mapping for the rotator cuff, methods must be refined for mapping thesupraspinatus tendon, and normative T2 values must first be acquired.
Materials and methods:
This study was IRB approved. Thirty asymptomatic volunteers (age: 18–62) wereevaluated with sagittal and coronal T2 mapping sequences. Manual segmentation of tendon and muscleas a unit and tendon alone was performed twice by two independent raters. Segmentations were dividedinto medial, middle and lateral subregions and mean T2 values calculated.
Results:
Anatomic comparison of mean T2 values illustrated highest values in the medial region, lowestvalues in the lateral region, and intermediate values for the middle region upon coronal segmentation(p < 0.001). In sagittal segmentations, there were higher values in the medial region and no signif-icant differences between the lateral and middle subregions. No significant differences were foundwith comparison across age groups. Inter and intra-rater segmentation repeatability was excellent, withcoefficients ranging from 0.85 to 0.99.
Conclusion:
T2 mapping illustrated anatomic variation along the supraspinatus muscle-tendon unit withlow standard deviations and excellent repeatability, suggesting that changes in structure due to degen-eration or changes associated with healing after repair may be detectable.
Adam W. Anza, Erin P. Lucasb, Eric K. Fitzcharlesb, Rachel K. Surowiecb,Peter J. Milletta, Charles P. Hob,∗
aThe Steadman Clinic, Vail, CO, United States
bSteadman Philippon Research Institute, Vail, CO, United States
Article history:
Received 26 July 2013
Received in revised form
16 December 2013
Accepted 3 February 2014
Keywords:
Rotator cuff
Supraspinatus
T2 mapping
MRI
Glenohumeral
1. Introduction
Evaluation of the rotator cuff has improved with advancementsin conventional MRI technology. Clinicians can gain insight intothe morphologic status of tendon to attachment with qualitativeinterpretation of images. However, the sensitivity and specificityof MRI to determine pathology has proven variable, especially inthe presence of tendonopathy and partial rotator cuff tears and theevaluation of signal intensity alone may contribute to lack of reli-ability [1–3]. Furthermore, diagnosis and monitoring of post-repairtendon health with MRI also remains largely subjective [4].The degenerative process of rotator cuff tendons involvesalterations in collagen structure and biochemical composition.Quantitative MRI has demonstrated to be sensitive to biochemi-cal changes in cartilage, recently in cartilage of the glenohumeraljoint, and could potentially be used to detect subtle changes in thebiochemical composition of tendons associated with tendon degen-eration. Initial studies have shown promise for detecting tendonpathology using quantitative MRI [5–10]. Pilot clinical applica-tion reported anatomic variation in T2* values within the Achillestendon-muscle unit in asymptomatic volunteers, and differencesin T2* values between asymptomatic volunteers and those withchronic Achilles tendinopathy [7]. To our knowledge, there are nopublished studies which analyze the rotator cuff tendons usingquantitative MRI mapping. Quantifiable, objective data on thehealth status of the rotator cuff would be exceptionally useful in aclinical setting for the non-invasive diagnosis of injury and trackingpost-surgical treatment response. To date, T2 mapping is the mostcommonly used mapping sequence within clinical practice, thusevaluation of T2 mapping for diagnosing rotator cuff pathology hashigh clinical relevance and potential for implementation.
Before applying T2 mapping scans to the clinical population, it is essential to appreciate normative T2 values and the distribution of these values within the rotator cuff tendons to serve as a baseline, similar to previous normative studies of the tendons utilizing classic MRI technology [11]. Therefore, the objective of this prospective study was evaluate verified asymptomatic volunteers using T2 mapping in order to establish a quantitative MRI protocol using T2 mapping for this purpose and to determine normative T2 valuesin asymptomatic tendons across age groups. We hypothesized that a normal range and pattern of T2 values exists which correlates to the anatomic structure of the muscle-tendon unit. Additionally,we hypothesized that due to compositional changes which occur during aging, differences in T2 values would be observed across age groups [12]. The results of this study will establish normative,asymptomatic T2 values for the supraspinatus tendon according to age, which can later be used as a reference and baseline for comparison for T2 mapping of damaged rotator cuff tendons and mapping of healing rotator cuff tendons after repair using these clinically relevant subregions.
2. Materials and methods
2.1. SubjectsThis study was approved by the Institutional Review Board and all subjects provided informed consent. Thirty recreationally active volunteers who were entirely asymptomatic for rotator cuff pathology of the shoulder were prospectively enrolled in the study. The volunteers were deemed asymptomatic by administering a short questionnaire regarding symptoms and injury history and by verifying a normal clinical examination that was performed by a sports medicine orthopedic surgeon. All volunteers were between theages of 18–62 (16 female, 15 male; 15 right shoulders, 15 leftshoulders) and were grouped into three equal groups with 15year age spans: 18–32 (5 female, 5 male; 5 left, 5 right), 33–47(4 female, 6 male; 5 left, 5 right), 48–62 (7 female, 3 male, 5left, 5 right). Grouping was determined by convenience sampling,as no prior tendon-based T2 mapping studies included similar comparisons that could be used for sample size guidance. Exclusion criteria included previous shoulder surgery or injury, history of glenohumeral joint disease and rotator cuff pathology found onexamination of subsequent clinical MRI. No distinction was made between dominant and non-dominant extremity; it was however paramount that the imaged side was asymptomatic. The time delay between physical examination and MR acquisition was limited to36 h to ensure that the volunteers remained asymptomatic. In addition, volunteers were asked to abstain from heavy lifting or exercise that would involve the upper body on the day of the scan in order to account for any changes in hydration levels due to physical activity.
2.2. Image acquisitionMR imaging was performed with a Siemens Magnetom Verio3.0T scanner (Siemens Medical Solutions, Erlangen, Germany) with a gradient strength of 40 mT/m, using a four channel small shoulder array coil (Invivo, Gainesville, FL, USA). Volunteers were positionedin the supine position with a sandbag placed in the supinated palm. The scanning protocol included an axial proton-density turbo-spin echo fat saturated sequence, followed by a coronal T2 mapping sequence, and finally a sagittal T2 mapping sequence. The T2mapping sequence was a multi-echo spin-echo (MESE) sequence(2000 ms repetition time; 10.7, 21.4, 32.1, 42.8, 53.5, 64.2, 74.9 msecho times; 140 mm field of view; 256 × 256 resolution; 2 mm slice thickness; 18 slices; 50% distance factor; 6:30 min scan time). The axial proton-density turbo-spin echo fat saturated clinical scan and the raw images from the sagittal and coronal MESE T2 mappingscans were reviewed to verify the absence of rotator cuff pathology. Clinical MRI findings of a partial rotator cuff tear or a full rotator cuff tear were grounds for exclusion. Values were calculated fromthe T2 mapping sequence using Siemens MapIt software (SiemensMedical Solutions, Erlangen, Germany).
Table 1
Inter and intra-rater ICC and confidence intervals.
| Inter-ratera | Intra-rater 1a | Intra-rater 2a | |||||
|---|---|---|---|---|---|---|---|
| Subregion name | Muscle tendon | Tendon | Muscle tendon | Tendon | Muscle tendon | Tendon | |
| Sagittal | Lateral | 0.87 [0.71,0.93] | 0.88 [0.76,0.94] | 0.89 [0.78,0.95] | 0.90 [0.80,0.95] | 0.94 [0.87,0.87] | 0.94 [0.88,0.97] |
| Middle | 0.90 [0.79,0.95] | 0.94 [0.88,0.97] | 0.94 [0.88,0.97] | 0.95 [0.90,0.98] | 0.97 [0.94,0.99] | 0.96 [0.92,0.98] | |
| Medial | 0.95 [0.88,0.97] | 0.74 [0.51,0.87] | 0.97 [0.94,0.99] | 0.84 [0.69,0.92] | 0.98 [0.96,0.99] | 0.88 [0.77,0.95] | |
| Coronal | Lateral | 0.93 [0.85,0.97] | 0.94 [0.88,0.97] | 0.98 [0.97,0.99] | 0.97 [0.93,0.98] | 0.97 [0.93,0.99] | 0.96 [0.93,0.98] |
| Middle | 0.95 [0.90,0.98] | 0.91 [0.81,0.96] | 0.98 [0.95,0.99] | 0.94 [0.88,0.97] | 0.95 [0.90,0.98] | 0.96 [0.92,0.98] | |
| Medial | 0.96 [0.91,0.98] | 0.86 [0.72,0.93] | 0.98 [0.95,0.99] | 0.92 [0.83,0.96] | 0.98 [0.96,0.99] | 0.90 [0.80,0.95] | |
a 95% CIs for inter-rater and intra-rater ICCs [LB, UB].
3. Results
All 30 volunteers were found to be asymptomatic for rotator cuff pathology upon questionnaire analysis and clinical exam. However, two of the 30 volunteers were excluded due to an evident full thickness supraspinatus tear found during the subsequent morphologic MRI evaluation. Both subjects had small isolated supraspinatus tendon tears with intrasubstance degeneration, and fatty infiltration and atrophy of the musculotendinous junction but no gross retraction of the tendon. For the 28 volunteers without tears, the inter-rater correlation coefficients and intra-rater correlation coefficients were all excellent except the medial subregion of the tendon alone segmentations which had fair-good repeatability (Table 1).In the coronal plane, mean T2 values of the muscle-tendon segmentations decreased significantly from medial to lateral for all three subregions (p < 0.001) (Fig. 2A). For the tendon segmentations, mean T2 values of the middle and lateral subregions were lower than the medial subregion (p < 0.001), though not significantly different from each other (p = 0.1) (Fig. 2B).In the sagittal plane, mean T2 values of the muscle-tendon segmentations were significantly higher in the medial region com-pared to the middle and lateral regions (p < 0.001), but no statistical difference was found between the middle and lateral regions(p = 1.0) (Fig. 2C). For the tendon segmentations, the lateral and medial subregions had higher T2 values than the middle region(p < 0.001), and these two regions were not significantly different from each other (p = 0.4) (Fig. 2D).When comparing the tendon and muscle-tendon segmentations to each other, the values for middle and medial subregions of the muscle-tendon segmentations were higher than the tendon segmentations (p < 0.001) and the lateral subregions were not significantly different (p = 1.0) in the sagittal plane. The same results were true for the coronal plane except the medial subregions were also not significantly different from each other (p = 0.7). No significant differences were found in mean T2 values between different age groups for either plane.4. Discussion
The present study applied T2 mapping to asymptomatic human rotator cuff tendons in vivo. T2 values in the rotator cuff tendons demonstrated significant variation based upon anatomic subregion, with higher values in the more medial subregions and lower values in the more lateral subregions. This variation is expected as the medial subregion contains more muscle than tendon, and thus has inherently higher water content. Structures with high water content have been shown to have increased T2 values compared with soft tissue structures of lower water content. These results are in support of the hypothesis in that T2 values would vary with varying tissue structure. The ability to detect significant differences across the subregions provides indication that T2 mapping may be useful in the clinical setting for detecting structural variation in the tendon and muscle.
It was unexpected that no significant differences were observed in T2 values between age groups since histological studies have shown changes in the biochemical composition of rotator cuff tendons with age [12]. A potential reason for these results was the strict inclusion criteria to ensure volunteers truly had asymptomatic and therefore likely healthier supraspinatus tendons. There were two subjects with asymptomatic tears, who were in the age group 51–66. A similar rate of asymptomatic tears has been documented in this age group, with a rate of 23% in a sample of subjects older than 50 years [14]. Review of the clinical MRI exams of both subjects in this cohort revealed full thickness tears with no gross retraction which caused the data from these subjects to be removed from the study. Although excluded for statistical analysis, the mean T2 values of the two volunteers with MRI findings of a supraspinatus tear were substantially higher than the standard deviations for all regions in the remaining asymptomatic volunteers except for the middle subregion (coronal plane). The mean T2 values from these volunteers are shown in Fig. 2.
T2 mapping scans were acquired in both the sagittal and coronal planes in order to determine an optimal plane for analysis and use in clinical practice. While the inter-rater and intra-rater correlation coefficients were mostly excellent for both planes indicating robustness of the manual segmentation and landmark selection,the raters felt more confident with segmenting the footprint of the tendon in the coronal plane. On the most lateral slices of the sagittalsequence, the raters reported difficulty separating tendon and bone at the footprint. This was due to the increase in tendon curvature out of the slicing plane and associated partial volume averaging and may have affected the T2 values in this region. For this reason, the authors believe the coronal sequence may be optimal for analyzing T2 values of the supraspinatus footprint. It should be noted that differentiation between the supraspinatus and infraspinatus is difficult in the coronal plane, because of partial volume averaging with the coronal orientation of the supraspinatus and infraspinatus. For investigations involving more medial regions or where muscle differentiation is imperative, the sagittal plane may be optimal since the entire musculotendinous cross section is easily visualized and can be reliably segmented.
The tendon and muscle-tendon segmentations were both obtained in order to evaluate the reliability of the segmentation method (i.e. tendon and muscle plus tendon combined). The inter-rater and intra-rater correlation coefficients were excellent for both segmentations, except for the medial subregion of the tendon segmentations which was fair-good. In general, the raters were more confident segmenting the muscle and tendon together in the medial subregions and noted that it was very difficult to distinguish the developing tendon and slips from the muscle about the myotendinous junction.
A limitation of the study is that biochemical validation of T2 mapping of the rotator cuff has not been performed to verify that variation in values has clinical meaning. Such validation would require histologic specimens in an animal or cadaveric human model. While the majority of research and clinical applications of quantitative MRI have focused on cartilage, initial efforts toward utilizing T2 mapping in the analysis of tendon health have been promising [6,15].
Asymptomatic subject data was collected in this study and thus we cannot conclude whether T2 mapping will be useful for detecting rotator cuff pathology. Results were promising, however, in that differences in T2 values were observed between subregions with known structural variation and were performed with good to excellent inter-rater and intra-rater correlation. However, analysis of the oldest age group may have been limited due to the exclusion of two volunteers with asymptomatic tears reducing the number of volunteers analyzed in this group. Future directions include evaluating patients with documented rotator cuff pathology to enable comparison. Ultimately, automated segmentation and automated landmark detection would be ideal for use of this technique in the clinical setting.
5. Conclusions
In conclusion, this study provides initial normative T2 values for asymptomatic supraspinatus tendons using clinically relevant subregions. Variability of T2 values between subregions is consistent with the muscle/tendon composition. Age was not a factor as it did not have a significant effect on T2 values in this asymptomatic cohort. The values from this study can be used for comparison to identify rotator cuff degeneration in patients presenting with abnormalities of the shoulder. Further studies are needed to analyze T2 mapping to document its clinical utility and value to improving assessment of rotator cuff pathology and post-repair healing.
Conflicts of interest
All work was performed at the Steadman Philippon Research Institute. The authors and Institute report no conflicts of interest pertaining to this work.
Acknowledgements
The authors would like to graciously acknowledge the statistical work of Grant Dornan, MSc and his contributions to this study. The authors would also like to acknowledge Bill Brock for acquiring the MRI scans used in the study.
References
[1] Balich SM, Sheley RC, Brown TR, Sauser DD, Quinn SF. MR imaging of the rotator cuff tendon: interobserver agreement and analysis of interpretive errors.Radiology 1997;204(1):191–4.[2] Reinus WR, Shady KL, Mirowitz SA, Totty WG. MR diagnosis of rotator cuff tears of the shoulder: value of using T2-weighted fat-saturated images. Am J Roentgenol 1995;164(6):1451–5.
[3] Robertson PL, Schweitzer ME, Mitchell DG, et al. Rotator cuff disorders: inter-observer and intraobserver variation in diagnosis with MR imaging. Radiology1995;194(3):831–5.
[4] Khazzam M, Kuhn JE, Mulligan E, et al. Magnetic Resonance Imaging Identification of Rotator Cuff Retears After Repair: Interobserver and Intraobserver Agreement. Am J Sports Med 2012;40:1722–7.
[5] Crema MD, Roemer FW, Marra MD, et al. Articular cartilage in the knee: cur-rent MR imaging techniques and applications in clinical practice and research.Radiographics 2011;31(1):37–61.
[6] Juras V, Apprich S, Pressl C, et al. Histological correlation of 7 T multi-parametric MRI performed in ex vivo Achilles tendon. Eur J Radiol 2011;82(5):740–4.
[7] Juras V, Zbyn S, Pressl C, et al. Regional variations of T(2)* in healthy and pathologic Achilles tendon in vivo at 7 Tesla: preliminary results. Magn Reson Med 2012;68(5):1607–13.
[8] Robson MD, Benjamin M, Gishen P, Bydder GM. Magnetic resonance imaging of the Achilles tendon using ultrashort TE (UTE) pulse sequences. Clin Radiol 2004;59(8):727–35.
[9] Trattnig S, Zbyn S, Schmitt B, et al. Advanced MR methods at ultra-high field (7 Tesla) for clinical musculoskeletal applications. Eur Radiol 2012;22(11):2338–46.
[10] Bittersohl B, Miese FR, Dekkers C, et al. T2* mapping and delayed gadolinium-enhanced magnetic resonance imaging in cartilage (dGEMRIC) of glenohumeral cartilage in asymptomatic volunteers at 3 T. Eur Radiol 2013;23(5):1367–74.
[11] Huber DJ, Sauter R, Mueller E, Requardt H, Weber H. MR imaging of the normal shoulder. Radiology 1986;158(2):405–8.
[12] Riley GP, Harrall RL, Constant CR, Chard MD, Cawston TE, Hazleman BL. Glycosaminoglycans of human rotator cuff tendons: changes with age and in chronic rotator cuff tendinitis. Ann Rheum Dis 1994;53(6):367–76.
[13] Fleiss JL. The design and analysis of clinical experiments. Wiley series in probability and mathematical statistics Applied probability and statistics. New York:Wiley; 1986.
[14] Tempelhof S, Rupp S, Seil R. Age-related prevalence of rotator cuff tears in asymptomatic shoulders. J Shoulder Elbow Surg 1999;8(4):296–9.
[15] Wellen J, Helmer KG, Grigg P, Sotak CH. Spatial characterization of T1 and T2 relaxation times and the water apparent diffusion coefficient in rabbit Achilles tendon subjected to tensile loading. Magn Reson Med 2005;53(3):535–44.
* Corresponding author at: Steadman Philippon Research Institute, 181 W. Meadow Dr, Suite 1000, Vail, CO 81657, United States. Tel.: +1 970 479 5888; fax: +1 970 479 9753.
E-mail addresses: anz.adam.w@gmail.com (A.W. Anz), erin.lucas14@gmail.com (E.P. Lucas), ericfitzcharles@gmail.com (E.K. Fitzcharles), Rachel.surowiec@sprivail.org (R.K. Surowiec), drmillett@thesteadmanclinic.com (P.J. Millett), Charles.ho@sprivail.org (C.P. Ho).
http://dx.doi.org/10.1016/j.ejrad.2014.02.002
0720-048X/© 2014 Elsevier Ireland Ltd. All rights reserved.
For additional information on ACL knee injuries, or to learn more about what is involved during ACL reconstruction surgery, please contact the office of orthopedic knee surgeon, Dr. Adam Anz, serving the greater Pensacola, Gulf Breeze, and Gulf Coast communities.
