University of Southern Denmark

Evidence-based research on effectiveness of periodontal treatment in rheumatoid arthritis

patients
a systematic review and meta-analysis
Silva, Daniela S; Costa, Flávio; Baptista, Isabel P; Santiago, Tânia; Lund, Hans; Tarp, Simon;
Silva, José A P; Christensen, Robin

Published in:
Arthritis Care & Research

DOI:
10.1002/acr.24622

Publication date:
2022

Document version:
Accepted manuscript

Citation for pulished version (APA):

Silva, D. S., Costa, F., Baptista, I. P., Santiago, T., Lund, H., Tarp, S., Silva, J. A. P., & Christensen, R. (2022).
Evidence-based research on effectiveness of periodontal treatment in rheumatoid arthritis patients: a systematic
review and meta-analysis. Arthritis Care & Research, 74(10), 1723-1735. https://doi.org/10.1002/acr.24622

Go to publication entry in University of Southern Denmark's Research Portal

Terms of use
This work is brought to you by the University of Southern Denmark.
Unless otherwise specified it has been shared according to the terms for self-archiving.
If no other license is stated, these terms apply:

• You may download this work for personal use only.

• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying this open access version
If you believe that this document breaches copyright please contact us providing details and we will investigate your claim.
Please direct all enquiries to [email protected]

Download date: 12. nov.. 2022

 RESEARCH ARTICLE

Biomechanical Effects of Chronic Ankle

Instability on the Talar Cartilage Matrix:
The Value of T1ρ Relaxation Mapping
Without and With Mechanical Loading
Thomas Lange, PhD,1* Lukas Sturm, MD,2 Pia M. Jungmann, MD,2 Matthias Jung, MD,2
Spartak Ovsepyan, MD,3 Marco Reisert, PhD,1,4 Hagen Schmal, MD,3,5 and
Markus Wenning, MD3

Background: T1ρ mapping has been proposed for the detection of early cartilage degeneration associated with chronic
ankle instability (CAI). However, there are limited data surrounding the influence of ankle loading on T1ρ relaxation.
Purpose: To evaluate T1ρ relaxation times of talar cartilage, as an indicator of early degenerative changes, associated with
CAI and to investigate the influence of acute axial in situ loading on T1ρ values in CAI patients and healthy controls.
Study Type: Prospective.
Subjects: A total of 9 patients (age = 21.8
2.5 years, male/female = 2/7) with chronic ankle instability and 18 healthy
control subjects (age = 22.8
3.6 years, male/female = 5/13).
Field Strength: 3 T.
Sequence: 3D gradient echo fast low-angle shot (FLASH) sequence augmented with a variable spin-lock preparation
period.
Assessment: Ankle T1ρ mapping was performed without and with axial loading of 500 N. The talar cartilage was seg-
mented in five coronal slices covering the central talocrural joint. Median talar T1ρ values were separately calculated for
the medial and lateral facets.
Statistical Tests: Mann–Whitney U and Wilcoxon signed-rank tests, significance level: P < 0.05.
Results: For the combined cohorts, the statistical analysis yielded significantly lower T1ρ values with loading compared to
the no-load measurement for both the lateral (no load: [51.0
4.0] msec, load: [49.5
5.4] msec) as well as the medial
compartment (no load: [50.0
5.4] msec, load: [47.8
6.8] msec). In the unloaded scans, the CAI patients showed signifi-
cantly increased talar T1ρ values ([53.0
7.4] mse ) compared to the healthy control subjects ([48.8
4.1] msec) in the
medial compartment.
Data Conclusion: Increased talar T1ρ relaxation times in CAI patients compared to healthy controls suggest that T1ρ relax-
ation is a sensitive biomarker for CAI-induced early-stage cartilage degeneration. However, the load-induced T1ρ change
did not prove to be a viable marker for the altered biomechanical properties of the hyaline talar cartilage.
Level of Evidence: 2
Level of Efficacy: Stage 2.
J. MAGN. RESON. IMAGING 2022.

View this article online at wileyonlinelibrary.com. DOI: 10.1002/jmri.28267

Received Feb 8, 2022, Accepted for publication May 11, 2022.

*Address reprint requests to: Thomas Lange, Medical Physics, Department of Radiology, Medical Center – University of Freiburg, Killianstrasse 5a, 79106 Frei-
burg, Germany. E-mail: [email protected]
Thomas Lange and Lukas Sturm contributed equally to this work.

From the 1Division of Medical Physics, Department of Diagnostic and Interventional Radiology, Medical Center – University of Freiburg, Faculty of Medicine,
University of Freiburg, Freiburg, Germany; 2Department of Diagnostic and Interventional Radiology, Medical Center – University of Freiburg, Faculty of
Medicine, University of Freiburg, Freiburg, Germany; 3Department of Orthopedic and Trauma Surgery, Medical Center – University of Freiburg, Faculty of
Medicine, University of Freiburg, Freiburg, Germany; 4Department of Stereotactic and Functional Neurosurgery, Medical Center – University of Freiburg,
Faculty of Medicine, University of Freiburg, Freiburg, Germany; and 5Department of Orthopaedic Surgery, Odense University Hospital, Odense, Denmark
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction
in any medium, provided the original work is properly cited and is not used for commercial purposes.

© 2022 The Authors. Journal of Magnetic Resonance Imaging published by Wiley Periodicals LLC 1
on behalf of International Society for Magnetic Resonance in Medicine.
Journal of Magnetic Resonance Imaging

A cute ankle sprain is one of the most common sports

injuries of the lower extremities.1 Approximately 20%–
40% of affected individuals develop chronic ankle instability
studies, there are limited data on T1ρ mapping in the
literature.16,17 In both of these studies, a CAI-associated T1ρ
increase has been reported.
(CAI).2 Depending on whether deficits are predominantly MRI markers such as T2 and T1ρ relaxation do not
mechanical or functional, two etiologies of CAI can be distin- only reflect cartilage composition and microstructure, but also
guished: functional ankle instability (FAI) and mechanical cartilage function, that is, the biomechanical properties.18
ankle instability (MAI).3 Especially the latter is associated Therefore, the response of cartilage T2 and T1ρ to mechani-
with focal cartilage overload due to an imbalanced load distri- cal loading has been investigated in numerous knee MRI
bution.4 This focal cartilage stress may lead to microstructural studies. Studies have examined load-induced T2 and T1ρ
damage of the cartilage matrix, eventually resulting in changes through MRI measurements performed before and
osteoarthritis.5 after exercise or have performed T2 and T1ρ mapping with
While plain radiography only allows diagnosing osteoar- in situ loading, mostly demonstrating a load-induced decrease
thritis at an advanced stage by depicting degenerative changes of the relaxation constants, which is primarily attributed to a
such as joint space narrowing or osteophytes, MRI enables reduction of the water content under loading.19–23 Compara-
the noninvasive evaluation of the articular cartilage based on tive T2 and T1ρ mapping of patellofemoral cartilage under in
signal intensity changes and morphologic alterations.6 How- situ loading has demonstrated T1ρ mapping to be slightly
ever, conventional MRI is limited regarding the quantitative more robust and reproducible than T2 mapping, mainly due
assessment of early-stage osteoarthritis.7 For detection of the to the more homogeneous contrast across different cartilage
subtle biochemical changes associated with early-stage carti- layers.23 However, relaxometric data of the ankle under load-
lage degeneration, quantitative MRI techniques such as T2, ing are scarce.
T2* and T1ρ mapping, delayed gadolinium-enhanced MRI The aim of this study was 1) to evaluate T1ρ relaxation
of cartilage (dGEMRIC), glycosaminoglycan chemical times of talar cartilage as an indicator of early degenerative
exchange saturation transfer (gagCEST) or diffusion-weighted changes in CAI and 2) to investigate the influence of acute
imaging (DWI) have been proposed.8–11 T2 relaxation rates axial in situ loading on T1ρ relaxation times in CAI patients
are associated with the water content, collagen fiber orienta- and healthy controls.
tion, and anisotropy of cartilage while T1ρ relaxation is also
sensitive to the proteoglycan content.12 Since initial cartilage
Materials and Methods
matrix degeneration is mainly characterized by a depletion of
proteoglycans before cell swelling and disruption of the colla- Participants
This study was approved by the institutional review board (protocol
gen matrix occur, T1ρ mapping has been shown to be slightly
#118/19) and was prospectively registered at the German Clinical
more sensitive for detection of early cartilage degeneration
Trials Register (#DRKS00016356). It was conducted according to
than T2 mapping.13,14 At ultra-high field strengths, T2* the Declaration of Helsinki in its current form and all subjects
mapping has been proposed as an alternative to T2 mapping declared informed consent prior to participation. Nine CAI patients
that alleviates the problem of specific absorption rate limita- and 18 control subjects were recruited as subcohorts of a larger study
tions.8 The dGEMRIC technique is based on a substantial on mechanical ankle instability.24 Selection criteria were based on
reduction of the T1 relaxation constant, which is induced by literature commendations of the International Ankle Consortium.25
an intravenously administered gadolinium-based contrast Subjective instability was defined via recurrent feelings of “giving
agent.9 In contrast to T1ρ mapping, which is sensitive to way” for more than 1 year after injury and recurrent sprains. CAI-
both proteoglycan and collage content, dGEMRIC measure- related symptoms were assessed using the questionnaire-based Cum-
ments specifically reflect the proteoglycan concentration.8,9 berland Ankle Instability Tool (CAIT), using scores below 24 and
above 28 as inclusion criteria for the CAI group and for the control
Due to its invasiveness and side effects, dGEMRIC does not
group, respectively.26
find widespread application in clinical studies, though. The
gagCEST method is an emerging high-field technique for
investigation of the proteoglycan content.10 However, it relies MRI Acquisition
on sophisticated data processing and is very sensitive to mag- MRI experiments were performed on a Magnetom Trio 3 T system
(Siemens Healthineers, Germany), using an eight-channel multi-
netic field inhomogeneities.15 DWI has been proposed for
purpose coil (NORAS MRI products, Germany) for signal reception.
investigation of the collagen microstructure since fractional
Knee loading was realized with a custom-built MR-compatible pneu-
anisotropy can serve as a marker of early-stage cartilage degen-
matic loading device enabling accurate load adjustment in the range
eration, but DWI is hampered by long scan times and strong of 0–500 N from the console room.
sensitivity to subject motion.11 T1ρ mapping was performed with a 3D fast low-angle shot
Due to the thin cartilage layers and limited image reso- (FLASH) sequence augmented with a variable spin-lock preparation
lution, MRI relaxometry of the ankle is challenging. While period, using the following parameters: echo train duration =
T2 mapping has already been performed in numerous CAI 2380 msec, repetition time (TR) = 9.3 msec, echo time

2
 Lange et al.: T1ρ Mapping in Chronic Ankle Instability

FIGURE 1: T1ρ mapping sequence consisting of a spin-lock preparation module with variable spin-lock duration τ, fat suppression
based on frequency-selective excitation and a repeated 3D FLASH module.

(TE) = 4.62 msec, field of view (FOV) = 80 mm, matrix

size = 256  256, in-plane resolution = 0.3 mm, 12 3D partitions,
slice thickness = 2.4 mm, readout bandwidth (BW) = 330 Hz/pixel,
readout along RL direction, scan duration = 5:00 minutes. T1ρ
encoding was performed with a spin-lock frequency of 500 Hz and five
different spin-lock durations (τ = 0/10/20/30/40 msec). The TRshot
between spin-lock preparations was set to 5 seconds. To render the
spin-lock preparation insensitive with respect to field inhomogeneities
(ΔB0), the preparation interval τ was split into two periods with oppo-
site spin-lock B1 phase, separated by a 180 pulse for ΔB0 refocusing
(90 xτ/2y180 yτ/2-y90 -x).27 Fat saturation was included
between the spin-lock preparation and the FLASH module. The
sequence diagram is shown in Fig. 1.
For the experiments, the subject was attached to the scanner
bed in a supine position with a weight lifting belt to prevent
loading-induced motion (Fig. 2). The foot of the loaded leg was
placed in the foot holder of the loading device with the ankle in neu-
tral position. The lower leg was strapped to a shank holder to keep
FIGURE 2: Experimental setup: The subject is attached to the
the knee in full extension during the measurement. T1ρ measure- scanner bed with a weight lifting belt. The subject’s foot is
ments were performed first without and then with an axial load of affixed to the foot holder of the pneumatic ankle loading device
500 N. A load of 500 N (corresponding to 50 kg) was chosen since and the lower leg is strapped to a shank holder for stabilization.
it is assumed to induce a substantial loading effect but does not MRI acquisitions are performed with an eight-channel
multipurpose coil consisting of two elements that are attached
exceed the weight on the ankle during one-legged standing. MRI
to the ankle from both sides.
scans with in situ loading were executed with a delay of at least
30 seconds after the onset of loading.

where S(τ) is the measured signal intensity of the image for a partic-
ular spin-lock duration τ and S0 is the signal intensity for τ = 0.
Postprocessing
From 12 reconstructed coronal T1ρ slices, 5 slices covering
Postprocessing and evaluation of the T1ρ maps were performed
the central talocrural joint region were selected for further evalua-
using a web-based framework for medical image analysis (Nora Med-
tion. To this end, a central slice at the apex of the talus was identi-
ical Imaging Platform, Freiburg, Germany). Datasets with excessive
fied and additionally the two contiguous slices in the anterior
motion artifacts were excluded from the analysis by a radiologist
direction as well as the two contiguous slices in the posterior direc-
with 7 years of experience (L.S.).
tion were included in the analysis. In all five evaluated T1ρ mapping
T1ρ maps were calculated with a monoexponential pixel-by-
slices, two investigators (L.S., 7 years of experience; S.O., 2 years of
pixel fit according to the following signal equation:
experience) manually segmented the talar articular cartilage for all
participants. Care was taken to include neither bone nor synovial
S ðτÞ ¼ S 0  expðτ=T1ρÞ ð1Þ
fluid in the regions of interest (ROIs) since this would strongly bias

3
Journal of Magnetic Resonance Imaging

TABLE 1. Demographics and Cumberland Ankle Instability Tool (CAIT) Scores of the Two Investigated Cohorts

Control CAI Mann–Whitney U

Group size 18 9
Sex (m/f) 5/13 2/7
Age (years) 22.8
3.6 21.8
2.5 P = 0.38
Height (cm) 171.9
9.9 173.4
8.2 P = 0.96
Weight (kg) 66.8
13.5 69.8
14.6 P = 0.78
CAIT score 29.8
0.6 17.6
4.6 P < 0.05

The P values obtained with a Mann–Whitney U test comparing the two cohorts are listed in the last column.
CAI = chronic ankle instability; m = male; f = female.

the results. Questionable voxels at the edge of the cartilage layer were Results
excluded from the analysis. The segmented talar cartilage region Demographics and CAIT scores of the two cohorts are listed
extended to both tibiotalar and fibulotalar contact areas. The region
in Table 1. All subjects tolerated the applied ankle load.
was subdivided into a lateral and a medial ROI at the midpoint
However, in one of the nine CAI patients and three of the
between the edges of the talar dome defined by maximum talar sur-
face curvature. T1ρ datasets from five subjects were analysed by both
18 control subjects, the loading gave rise to motion artifacts
investigators to evaluate the interrater reliability via the Pearson cor- in the acquired MRI images, which impeded the segmenta-
relation coefficient and one subject was analysed three times to assess tion of the talar cartilage. For these four subjects, no quantifi-
the repeatability of the ROI analysis. cation of T1ρ under loading was performed. Therefore, only
8 CAI patients and 15 control subjects were included in the
statistical analyses involving loaded measurements. Exemplary
Statistical Analysis coronal MRI slices without (τ = 0) and with maximum T1ρ
Median T1ρ values were calculated for each 3D ROI consisting of contrast (τ = 40 msec) from unloaded and loaded scans on a
the five selected slices. Accounting for the small cohort sizes, differ- healthy control subject and a CAI patient are displayed in
ent groups (CAI vs. HC) and conditions (load vs. no-load) were Fig. 3. Figure 4 shows exemplary coronal T1ρ maps from a
compared using Mann–Whitney U and Wilcoxon signed-rank tests, healthy control subject without and with axial loading, illus-
respectively, and a significance threshold of P = 0.05, without trating a load-induced T1ρ decrease in the talar cartilage.
adjustments for multiple comparisons. Due to the small voxel size and thus limited signal-to-noise

FIGURE 3: Exemplary images from the T1ρ mapping acquisitions without (τ = 0) and with maximum T1ρ contrast (τ = 40 msec).
Coronal slices covering the apex of the talus are shown for unloaded and loaded measurements on a healthy control subject (male,
age: 20 years, height: 176 cm, weight: 70 kg) and a CAI patient (female, age: 22 years, height: 164 cm, weight: 75 kg).

4
 Lange et al.: T1ρ Mapping in Chronic Ankle Instability

FIGURE 4: Coronal slices (from the apex of the talus) of T1ρ datasets acquired from a healthy control subject (male, age: 20 years,
height: 176 cm, weight: 70 kg) without (top) and with (bottom) axial loading of 500 N. Left: First contrast of the T1ρ protocol
without any spin lock encoding. Lateral and medial regions of interest as evaluated in the study are shown in blue and red color,
respectively. Right: Corresponding color-coded T1ρ maps in ms.

ratio, a large T1ρ variance across the investigated cartilage compared to the no-load measurement for both the lateral
regions was observed. T1ρ histograms showed a strongly (no load: [51.0
4.0] msec, load: [49.5
5.4] msec) as well
right-skewed distribution because on the lower end T1ρ out- as the medial compartment (no load: [50.0
5.4] msec,
liers were limited by zero while on the higher end single load: [47.8
6.8] msec). However, the significance vanished
pixels reached values of several hundred milliseconds. To in both compartments when the two cohorts were analyzed
avoid a bias toward erroneously high T1ρ values, median separately. A comparison of the two cohorts yielded higher
T1ρ values instead of mean values are therefore reported. T1ρ values for the CAI patients compared to the healthy con-
Assessment of the interrater reliability yielded Pearson correla- trol subjects in the no-load measurements (Fig. 6). For the
tion coefficients of r = 0.99 (lateral ROI) and r = 0.99 medial compartment (control: [48.8
4.1] msec, CAI:
(medial ROI) for the unloaded scans, and r = 0.98 (lateral [53.0
7.4] msec), this effect was significant while no signif-
ROI) and r = 0.94 (medial ROI) for the loaded scans. icance was observed for the lateral compartment (control:
Figure 5 illustrates the repeatability of the T1ρ analysis as [50.1
4.3] msec, CAI: [53.0
3.7] msec, P = 0.07).
estimated from three repeated cartilage segmentations per- When comparing T1ρ of CAI and control subjects under
formed by the same investigator (S.O.) on a healthy subject. loading, no significant differences were found, neither for the
A statistical analysis on the combined cohorts including all lateral (control: [48.9
4.2] msec, CAI: [50.7
7.3] msec,
subjects yielded significantly lower T1ρ values with loading P = 0.19) nor for the medial talar cartilage (control:
[46.8
5.4] msec, CAI: [49.7
8.9] msec, P = 0.19).
Finally, a group comparison of the load-induced T1ρ changes
(Fig. 6) did not yield any significant differences between CAI
and control subjects (lateral compartment: control:
[1.2
2.4] msec, CAI: [1.5
3.0] msec, P = 0.43);
medial compartment: control: [2.0
4.3] msec, CAI:
[2.2
4.1] msec, P = 0.73). A summary of the statistical
evaluation can be found in Table 2.

Discussion
Few MRI studies have investigated the role of T1ρ relaxation
in CAI.16,17 The observation of increased T1ρ times in the
FIGURE 5: Repeatability of the T1ρ analysis as estimated from talar cartilage of CAI patients is in line with previous CAI
three repeated cartilage segmentations performed by the same
studies even though absolute values are difficult to compare
investigator (S.O.) on a healthy subject. Median T1ρ values
(in ms) are plotted for the medial and lateral ROIs without and due to different T1ρ acquisition and postprocessing proto-
with loading. cols.16,17 Since a total scan duration of 12 or 18 minutes as

5
Journal of Magnetic Resonance Imaging

FIGURE 6: Box plots showing the T1ρ statistics for the two subject cohorts (CAI and control) in the two regions of interest (top:
lateral, bottom: medial). Left: Median T1ρ values across ROI without ankle loading. Right: T1ρ change induced by axial loading of
500 N. The p values as obtained with a Mann–Whitney U test comparing the two cohorts are listed at the bottom (with an asterisk
denoting significance).

used in these previous ankle studies16,17 was not compatible patients, again with stronger effects in the medial joint
with the strong in situ loading applied in this work, consider- regions.31–33 Indeed, cartilage lesions have predominantly
ably longer echo train durations had to be used in the been found in the medial joint compartment in arthroscopy34
FLASH module (2380 msec compared to 443 msec). This and cadaver studies.35 While acute sprains may result in trau-
may give rise to increasing T1 contamination of the resulting matic cartilage injury on the lateral side, chronic instability
T1ρ contrast for the acquisition of outer k-space regions, predominantly gives rise to degeneration on the medial talar
which are sampled toward the end of the echo train, and con- shoulder. Furthermore, biomechanical cadaver experiments
sequently also to T1-induced image blurring. This effect have confirmed that the stress distribution is shifted towards
might be mitigated by using magnetization-prepared angle- the medial contact area of the talocrural joint after sectioning
modulated partitioned k-space spoiled gradient echo snap- the lateral ankle ligaments.36 Thus, the clinical appearance of
shots (MAPSS).28 It should be noted that T1ρ relaxation also CAI is adequately represented by T1ρ examination.
strongly depends on the applied spin-lock frequency. For A T1ρ protocol was chosen for this relaxometric study
most efficient T1ρ relaxation, the spin-lock frequency should since in a previous work on comparative T2 and T1ρ map-
ideally match the proton chemical exchange rates of the ping of the patellofemoral cartilage with in situ loading, T1ρ
hydroxyl and amide functional groups of the proteoglycans could be more reproducibly measured.23 It must be noted
(1500 Hz).29 However, due to specific absorption rate limi- that the magic angle effect has a major influence on T2 and
tations for in vivo examinations, a spin-lock frequency of T1ρ relaxation in articular cartilage so that relaxometric data
500 Hz was chosen in accordance with previous stud- strongly depend on fiber orientation and anisotropy.37,38
ies.16,17,30 While a significant difference was found for the However, this effect is less pronounced for T1ρ than for T2
medial joint compartment, only a nonsignificant trend was relaxation.37 While the sensitivity to fiber anisotropy benefits
observed for the lateral compartment in the unloaded mea- the detection of cartilage degeneration, which is usually asso-
surements, suggesting that the medial talar cartilage is more ciated with a loss of anisotropy, the sensitivity to fiber orien-
sensitive to CAI-induced microstructural changes. These T1ρ tation may encumber studies with comparative relaxometric
findings are also in agreement with several ankle studies that measurements, for example, longitudinal studies. As ankle
have reported a T2 increase in the talocrural cartilage of CAI loading can also have an influence on joint orientation, this

6
 Lange et al.: T1ρ Mapping in Chronic Ankle Instability

TABLE 2. Comparison of T1ρ Quantification Results Between Groups and Conditions (With the Group Sizes n Given
in Brackets)

Group Condition Lateral Medial

All Unloaded (n = 23) Mean 51.0 50.0
SD 4.0 5.4
All loaded (n = 23) Mean 49.5 47.8
SD 5.4 6.8
Δ 1.5 2.2
P <0.05 <0.05
Control Unloaded (n = 18) Mean 50.1 48.8
SD 4.3 4.1
CAI Unloaded (n = 9) Mean 53.0 53.0
SD 3.7 7.4
Δ 2.9 4.1
P 0.07 <0.05
Control Loaded (n = 15) Mean 48.9 46.8
SD 4.2 5.4
CAI Loaded (n = 8) Mean 50.7 49.7
SD 7.3 8.9
Δ 1.8 2.9
P 0.19 0.19
Control Δ (loaded–unloaded) (n = 15) Mean 1.2 2.0
SD 2.4 4.3
CAI Δ (loaded–unloaded) (n = 8) Mean 1.5 2.2
SD 3.0 4.1
P 0.43 0.73

The first table (orange) shows a T1ρ comparison of unloaded vs. loaded cartilage across both cohorts. The second and third table com-
pare results between the two cohorts (CAI and control) for the unloaded (green) and loaded (blue) condition. The last table (gray) repre-
sents a comparison of the load-induced T1ρ changes between the two cohorts. The P values of statistical tests as described in the
Methods section are reported (bold values indicate statistical significance: P < 0.05).

could give rise to an undesired bias when data acquired with measured load-induced T1ρ change was not significantly dif-
and without loading are compared. ferent between the CAI and control group and thus could
A significant load-induced T1ρ decrease was observed not be established as a suitable biomarker for early CAI-
for the combined cohorts in both the medial and lateral joint induced cartilage degeneration. In contrast, Souza et al22 have
compartments. This observation is in line with knee cartilage reported a significantly larger load-induced T1ρ decrease in
T1ρ studies involving in situ loading.22,23 However, the sig- the tibiofemoral cartilage of subjects with osteoarthritis com-
nificance vanished when a separate statistical analysis was per- pared to healthy controls, suggesting that more advanced car-
formed for the two cohorts. This suggests that larger tilage degeneration may indeed be reflected by an altered T1ρ
cohorts—at least of the size of the combined cohorts behavior under loading. It should be noted that for the sub-
(n = 23)—are required for robust detection of load-induced jects investigated in our study 3D turbo-spin echo (TSE) data
T1ρ changes. In this context, it is not suprising that the were concomitantly acquired in the context of a larger CAI

7
Journal of Magnetic Resonance Imaging

study published elsewhere.24 In the TSE images from the mainly occurs during the first 20 seconds of loading and
CAI patients, no higher-grade cartilage defects were identi- remains relatively constant after 50 seconds.40 We chose a
fied. A previous study performed in patients with functional delay of 30 seconds to ensure that on the one hand most of
ankle instability (FAI) has investigated tibiotalar T2 alter- the compression dynamics happens before the MRI measure-
ations between early-unloading and late-unloading.33 The ment and on the other hand the loading period is not sub-
FAI group of that study showed a T2 increase in the dorso- stantially prolonged and thus still tolerated by the subjects.
medial joint compartment after prolonged unloading while a
T2 decrease was observed in the antero-lateral aspect. In con- Conclusion
trast, T2 was increased across the whole tibiotalar joint in a Increased T1ρ relaxation times were observed in the talar car-
coper group (subjects who had suffered an ankle sprain but tilage of CAI patients compared to healthy controls, with the
did not experience residual CAI symptoms) after extended highest values in the medial joint compartment, suggesting
unloading, and no substantial differences were observed for a that T1ρ relaxation is a sensitive biomarker for CAI-induced
healthy control group. This observation was attributed to an early-stage cartilage degeneration. Furthermore, the T1ρ of
intercartilage hydration shift in FAI patients due to unbal- talar cartilage significantly decreased under in situ loading, in
anced ankle loading. Our findings may diverge from these accordance with T1ρ knee studies from the literature. How-
observations due to different loading paradigms. The static ever, in this exploratory pilot study, the load-induced T1ρ
axial in situ loading performed in our work gives rise to a change could not be established as a sensitive marker for the
more even and controlled load distribution than the dynamic altered biomechanics associated with CAI. Larger cohort sizes
loading associated with walking. However, the relaxometric may be required to elucidate this issue.
behavior under loading might provide insight into the ankle
biomechanics of patients with more advanced tibiotalar carti-
lage degeneration or other ankle pathologies such as Acknowledgments
osteochondrosis dissecans. The authors thank Dr. Ari Borthakur and Dr. Walter
Witschey from the University of Pennsylvania for providing
Limitations the T1ρ mapping sequence. This work was funded in part by
First, it should be noted that this is a pilot study with small the German Research Foundation (DFG), grant number LA
and uneven group sizes and results will have to be verified in 3353/4-1 and in part by the Bauerfeind AG, Zeulenroda,
larger cohorts. Second, the most anterior and posterior parts Germany. The sponsor had no influence on the study’s con-
of the talar cartilage were excluded from the analysis, and the tent, results or interpretation. There is no personal or any
tibial cartilage was not evaluated at all. Third, robust separate other relation which may pose a conflict of interest for any of
evaluation of deep and superficial cartilage layers as often con- the authors contributing to this publication. M.W. and
ducted in knee cartilage studies22,23 and also in some T2 M.J. were supported by the Berta-Ottenstein-Programme for
ankle studies from the literature31,33 was not feasible due to Clinician Scientists, Faculty of Medicine, University of Frei-
the small talar cartilage thickness and the limited spatial reso- burg. P.M.J. was supported by the Berta-Ottenstein-
lution of the acquired MRI data. Unfortunately, an extension Programme for Advanced Clinician Scientists, Faculty of
of the investigated cartilage regions and an enhanced spatial Medicine, University of Freiburg. Open Access funding
resolution could not be realized since many CAI patients enabled and organized by Projekt DEAL.
would not have tolerated a substantially prolonged loading
period. Fourth, it must be noted that in situ loading gives rise References
to a slight displacement of the MRI slab, which hampers the 1. de Noronha M, Refshauge KM, Herbert RD, Kilbreath SL. Do voluntary
comparative assessment of loaded and unloaded scans. There- strength, proprioception, range of motion, or postural sway predict
occurrence of lateral ankle sprain? Br J Sports Med 2006;40:824-828.
fore, the cartilage regions evaluated with and without loading
are not entirely identical. Prospective motion correction with 2. Doherty C, Bleakley C, Hertel J, Caulfield B, Ryan J, Delahunt E. Recov-
ery from a first-time lateral ankle sprain and the predictors of chronic
interscan position locking as previously proposed for knee ankle instability: A prospective cohort analysis. Am J Sports Med 2016;
MRI with in situ loading23,39 might be beneficial in this 44:995-1003.

respect, but marker fixation around the ankle may represent a 3. Hertel J. Functional anatomy, pathomechanics, and pathophysiology of
lateral ankle instability. J Athl Train 2002;37:364-375.
major challenge. Fifth, subjects did not undergo a longer
period of unloading before the MRI examination while an 4. Caputo AM, Lee JY, Spritzer CE, et al. In vivo kinematics of the
tibiotalar joint after lateral ankle instability. Am J Sports Med 2009;37:
influence of the unloading period on tibiotalar T2 relaxation 2241-2248.
has been demonstrated.33 Furthermore, it should be noted
5. Valderrabano V, Hintermann B, Horisberger M, Fung TS. Ligamentous
that the delay of 30 seconds between the onset of loading and posttraumatic ankle osteoarthritis. Am J Sports Med 2006;34:612-620.
the start of the MRI scan represents a compromise. A recent 6. Marinetti A, Tessarolo F, Ventura L, et al. Morphological MRI of knee
fluoroscopy study suggests that knee cartilage deformation cartilage: Repeatability and reproducibility of damage evaluation and

8
 Lange et al.: T1ρ Mapping in Chronic Ankle Instability

correlation with gross pathology examination. Eur Radiol 2020;30: 24. Wenning M, Gehring D, Lange T, et al. Clinical evaluation of manual
3226-3235. stress testing, stress ultrasound and 3D stress MRI in chronic mechani-
cal ankle instability. BMC Musculoskelet Disord 2021;22:198.
7. Zhang M, Min Z, Rana N, Liu H. Accuracy of magnetic resonance imag-
ing in grading knee chondral defects. Arthrosc J Arthrosc Relat Surg 25. Gribble PA, Delahunt E, Bleakley CM, et al. Selection criteria for
2013;29:349-356. patients with chronic ankle instability in controlled research: A position
statement of the international ankle consortium. J Athl Train 2014;49:
8. Schreiner MM, Mlynarik V, Zbýň Š, et al. New technology in imaging 121-127.
cartilage of the ankle. Cartilage 2017;8:31-41.
26. Hiller CE, Refshauge KM, Bundy AC, Herbert RD, Kilbreath SL. The
9. Burstein D, Velyvis J, Scott KT, et al. Protocol issues for delayed Cumberland ankle instability tool: A report of validity and reliability
Gd(DTPA)(2-)-enhanced MRI (dGEMRIC) for clinical evaluation of articu- testing. Arch Phys Med Rehabil 2006;87:1235-1241.
lar cartilage. Magn Reson Med 2001;45:36-41.
27. Witschey WRT, Borthakur A, Elliott MA, et al. Artifacts in T1ρ-weighted
10. Abrar DB, Schleich C, Radke KL, et al. Detection of early cartilage imaging: Compensation for B1 and B0 field imperfections. J Magn
degeneration in the tibiotalar joint using 3 T gagCEST imaging: A feasi- Reson 2007;186:75-85.
bility study. Magn Reson Mater Phys Biol Med 2021;34:249-260.
28. Li X, Han ET, Busse RF, Majumdar S. In vivoT1ρ mapping in cartilage
11. Raya JG. Techniques and applications of in vivo diffusion imaging of using 3D magnetization-prepared angle-modulated partitionedk-space
articular cartilage: Techniques of DTI of articular cartilage. J Magn spoiled gradient echo snapshots (3D MAPSS). Magn Reson Med 2008;
Reson Imaging 2015;41:1487-1504. 59:298-307.

12. Keenan KE, Besier TF, Pauly JM, et al. Prediction of glycosaminoglycan 29. Duvvuri U, Goldberg AD, Kranz JK, et al. Water magnetic relaxation
content in human cartilage by age, T1ρ and T2 MRI. Osteoarthr Cartil dispersion in biological systems: The contribution of proton exchange
2011;19:171-179. and implications for the noninvasive detection of cartilage degradation.
Proc Natl Acad Sci 2001;98:12479-12484.
13. Regatte RR, Akella SV, Lonner JH, Kneeland JB, Reddy R. T1rho relaxa-
tion mapping in human osteoarthritis (OA) cartilage: Comparison of 30. Witschey WR, Borthakur A, Fenty M, et al. T1rho MRI quantification of
T1rho with T2. J Magn Reson Imaging 2006;23:547-553. arthroscopically confirmed cartilage degeneration. Magn Reson Med
2010;63:1376-1382.
14. Sasho T, Katsuragi J, Yamaguchi S, et al. Associations of three-
dimensional T1 rho MR mapping and three-dimensional T2 mapping 31. Park SY, Yoon YC, Cha JG, Sung KS. T2 relaxation values of the Talar
with macroscopic and histologic grading as a biomarker for early articu- trochlear articular cartilage: Comparison between patients with lateral
lar degeneration of knee cartilage. Clin Rheumatol 2017;36:2109-2119. instability of the ankle joint and healthy volunteers. Am J Roentgenol
2016;206:136-143.
15. Ling W, Regatte RR, Navon G, Jerschow A. Assessment of glycosami-
noglycan concentration in vivo by chemical exchange-dependent satu- 32. Tao H, Hu Y, Qiao Y, et al. T2 -mapping evaluation of early cartilage
ration transfer (gagCEST). Proc Natl Acad Sci 2008;105:2266-2270. alteration of talus for chronic lateral ankle instability with isolated ante-
rior talofibular ligament tear or combined with calcaneofibular ligament
16. Wikstrom EA, Song K, Tennant JN, Dederer KM, Paranjape C, tear. J Magn Reson Imaging 2018;47:69-77.
Pietrosimone B. T1ρ MRI of the talar articular cartilage is increased in
33. Golditz T, Steib S, Pfeifer K, et al. Functional ankle instability as a risk
those with chronic ankle instability. Osteoarthr Cartil 2019;27:646-649.
factor for osteoarthritis: Using T2-mapping to analyze early cartilage
17. Song K, Pietrosimone B, Tennant JN, et al. Talar and subtalar T1ρ degeneration in the ankle joint of young athletes. Osteoarthr Cartil
relaxation times in limbs with and without chronic ankle instability. Car- 2014;22:1377-1385.
tilage 2021;13(1 Suppl):1402S-1410S.
34. Taga I, Shino K, Inoue M, Nakata K, Maeda A. Articular cartilage
18. Nieminen MT, Töyräs J, Laasanen MS, Silvennoinen J, Helminen HJ, lesions in ankles with lateral ligament injury: An arthroscopic study.
Jurvelin JS. Prediction of biomechanical properties of articular cartilage Am J Sports Med 1993;21:120-127.
with quantitative magnetic resonance imaging. J Biomech 2004;37:
35. Hirose K, Murakami G, Minowa T, Kura H, Yamashita T. Lateral liga-
321-328.
ment injury of the ankle and associated articular cartilage degeneration
19. Subburaj K, Kumar D, Souza RB, et al. The acute effect of running on in the talocrural joint: Anatomic study using elderly cadavers. J Orthop
knee articular cartilage and meniscus magnetic resonance relaxation Sci off J Jpn Orthop Assoc 2004;9:37-43.
times in young healthy adults. Am J Sports Med 2012;40:2134-2141. 36. Noguchi K. Biomechanical analysis for osteoarthritis of the ankle. Nihon
20. Nag D, Liney GP, Gillespie P, Sherman KP. Quantification of T2 relaxa- Seikeigeka Gakkai Zasshi 1985;59:215-222.
tion changes in articular cartilage with in situ mechanical loading of the 37. Shao H, Pauli C, Li S, et al. Magic angle effect plays a major role in
knee. J Magn Reson Imaging 2004;19:317-322. both T1rho and T2 relaxation in articular cartilage. Osteoarthr Cartil
2017;25:2022-2030.
21. Nishii T, Kuroda K, Matsuoka Y, Sahara T, Yoshikawa H. Change in
knee cartilage T2 in response to mechanical loading. J Magn Reson 38. Horiuchi S, Yu HJ, Luk A, et al. T1rho and T2 mapping of ankle carti-
Imaging 2008;28:175-180. lage of female and male ballet dancers. Acta Radiol Stockh Swed
1987;2020(61):1365-1376.
22. Souza RB, Kumar D, Calixto N, et al. Response of knee cartilage T1rho
and T2 relaxation times to in vivo mechanical loading in individuals 39. Lange T, Maclaren J, Herbst M, Lovell-Smith C, Izadpanah K,
with and without knee osteoarthritis. Osteoarthr Cartil 2014;22:1367- Zaitsev M. Knee cartilage MRI with in situ mechanical loading using
1376. prospective motion correction: Knee cartilage MRI using prospective
motion correction. Magn Reson Med 2014;71:516-523.
23. Lange T, Knowles BR, Herbst M, Izadpanah K, Zaitsev M. Comparative
T2 and T1ρ mapping of patellofemoral cartilage under in situ mechani- 40. Hosseini A, Van de Velde SK, Kozanek M, et al. In-vivo time-dependent
cal loading with prospective motion correction: T2 and T1ρ mapping of articular cartilage contact behavior of the tibiofemoral joint. Osteoarthr
knee cartilage. J Magn Reson Imaging 2017;46:452-460. Cartil 2010;18:909-916.