Int J Sports Med 2018; 39(01): 37-49
DOI: 10.1055/s-0043-118008
Orthopedics & Biomechanics
© Georg Thieme Verlag KG Stuttgart · New York

MRI Cartilage Assessment of the Subtalar and Midtarsal Joints During a Transcontinental Ultramarathon – New Insights into Human Locomotion

Uwe Hans-Werner Schütz1, Christian Billich1, Daniel Schoss1, Meinrad Beer1, Jutta Ellermann2
  • 1Universitatsklinikum Ulm, Klinik für Diagnostische und Interventionelle Radiologie, Ulm, Germany
  • 2University of Minnesota, Department of Radiology (CMRR), Minneapolis, United States
Further Information

Correspondence

Uwe Hans-Werner Schütz, PD Dr.med.
Universitatsklinikum Ulm
Klinik für Diagnostische und Interventionelle Radiologie
Albert-Einstein-Allee 29
89075 Ulm
Germany   
Phone: +49/171/7628 003   
Fax: +49/731/500 61002   

Publication History



accepted after revision 19 July 2017

Publication Date:
30 November 2017 (eFirst)

 

Abstract

MR measurements can be accurately performed during 4486 km of running, opening a window into in vivo assessment of hindfoot articular cartilage under extreme ultra-endurance loading. This observational cross-sectional study included 22 randomized participants of TransEurope FootRace between Italy and the North Cape, which was accompanied by a trailer-mounted 1.5T MRI scanner over 9 weeks. Four follow up MR examinations of subtalar and midtarsal joints were performed. Statistics of cartilage T2*  and thickness were obtained. Nearly all observed joints showed an initial significant mean T2*  increase of 20.9% and 26.3% for the left and right side, followed by a relative decrease of 28.5% and 16.0% during the second half, respectively. It could be demonstrated that mobile MRI field studies allow in vivo functional tissue observations under extreme loading. Elevated T2*  values recovered during the second half of the ultramarathon supported the evidence that this response is a physiological adaptive mechanism of chondrocyte function via upregulation of de novo synthesis of proteoglycans and collagen. These changes occurred in a distinct asymmetric pattern leaving a “biochemical signature” of articular cartilage that allows in vivo insight into joint loading. In conclusion, the normal articular cartilage of the hindfoot is resilient and adaptive, leaving extreme endurance activities up to limitless human ambition.


#

Introduction

The subtalar and midtarsal joints of the hindfoot act as an important unit for gait and function linking ankle and foot. These joints are difficult to conceptualize and study in vivo.

The subtalar joint is made up of the posterior, middle, and anterior facets of the talocalcaneal joint (TCJ) and acts synchronously with the ankle and midtarsal joints during gait. The midtarsal joint (MTJ), also known as transverse tarsal or Chopart joint, consists of the talonavicular (TNJ) and calcaneocuboid (CCJ) joints and plays a critical role in allowing the foot to transition from a flexible structure that dissipates impact of the body’s weight as the foot strikes the ground to the rigid structure that is required for efficient propulsion during takeoff [36]. Current information obtained from studies of human locomotion on the subtalar and midtarsal joints is based on external measurements, such as alignment and ground reaction forces, and lacks information on the “internal response,” most importantly, the response of the articular cartilage exposed to repetitive biomechanical loading during endurance running.

Results of this study are part of the interdisciplinary MRI-based TransEurope FootRace (TEFR) project. This field study was realized on the second multistage ultramarathon TEFR from the south (Bari, Italy) to the northernmost border of Europe (North Cape, Norway) and is the first and unique investigation showing feasibility of accompanying and continuously monitoring the participants with a mobile MRI travelling along on a semitrailer truck day by day from stage to stage over more than 9 weeks [41].

Limited information is available about the impact of high-level endurance running on lower extremity joints [23]. Studies concerning the question if long-distance running can cause osteoarthritis come to contradictory conclusions [6] [8] [25]. Although sparse data are available on the ankle cartilage and its functional adaptation to mechanical loading [38] [44], only the TEFR project focused on ultralong marathon impact on the ankle joint [12], and until now, no data are published regarding hindfoot cartilage in general regarding running and in special regarding ultralong endurance running.

Magnetic resonance imaging (MRI) is a reliable method for quantitative and qualitative assessment of human cartilage joint morphology [9] [45]. Joint articular cartilage volume and thickness measurement methods are precise and reproducible as previously demonstrated in the knee [10] [11] [20]. Quantitative MRI relaxation time measurements, such as T2 or T2*  of articular cartilage have been shown to correlate with interstitial fluid bound to proteoglycan molecules [14] as well as the equilibrium of intrachondral free water and in the collagen fiber orientation [29] which is influenced by physiological states related to rest and loading.

We hypothesize that T2*  measurements of the articular cartilage can be accurately performed at the subtalar and midtarsal joints, opening a window into “internal” functional assessment of human locomotion. Furthermore, we hypothesize that quantitative MR T2*  changes over time and when compared with baseline allow for in vivo assessment of the effect of joint loading during 4486 km (2787.5 miles) of ultralong distance running on the articular cartilage of bilateral subtalar and midtarsal joints.


#

Materials and Methods

Study participants

The local ethics committee approved the prospective study in accordance with international standards and the Declaration of Helsinki (Ref.No. 08-UBB/se) [16] [41]. Data were handled in accordance with HIPPA guidelines. The following inclusion criteria were applied: a) official race participant, b) minimum age of 18 years, c) medical health certificate and d) proof of appropriate ultramarathon running performance. Data were acquired between April and June 2009 and 66 volunteers were screened for MRI contraindications. Forty-four volunteers (67% of the race participants) were enrolled in the entire MRI study after verbal and written informed consent was obtained, and 22 participants were cluster-randomized for the MRI study of the subtalar and midtarsal joints [41]. Thirteen of them (59.1%) completed the race. Premature race termination was caused by soft tissue overuse symptoms, all unrelated to the joints (6). Data of the study volunteers (13 finishers/9 nonfinishers) at baseline were: mean age 45.4/53.8 years (SD 10.7/11.3, range 27–62/36–68), male 12/8 (92.7%/88.9%), BM 73.0/67.9 kg (SD 11.3/11.3, range 51.9–94.2/49.2–81.8), BMI 23.4/23.5 kg/m2 (SD 2.5/3.0, range 20.5–29.1/19.2–28.3). This study is embedded in a comprehensive project to improve our understanding of the physiological effects of ultraendurance on body mass and fat index [12], the osseous and muscular structures of the foot [39] and the articular cartilage of the ankle joint [40] with MRI data acquired from the same volunteer group during TEFR. Our in vivo results on the subtalar (talocalcaneal) and midtarsal (talonavicular and calcaneocuboid) articular cartilage are uniquely different and novel, because there is no prior publication on in vivo biochemical cartilage assessment under endurance loading of these joints.


#

Data acquisition

MRI data were acquired with a trailer-mounted 1.5T MR scanner (AvantoTM mobile MRI 02.05, SyngoTM; gradient strength of 45 mT/m, Siemens Ltd., Germany) travelling with the runners throughout the entire duration of TEFR [2]. A dedicated, eight-channel foot and ankle coil was utilized. Scanning took place at baseline (t0) within four days before the start of the race and at about every 900 km ( +/– 211.5 km), between 2:00 pm and 9:00 pm after the daily stage (t1–t4) during TEFR ([Table 1]), with t4 representing the final measurement. The initial data points (t1) were obtained after a mean distance run of 1200 km (range 1068 to 1457 km, stage 16–22), during which the effects of long-distance running prevailed. For T2*  mapping, a sagittal GRE (gradient echo) sequence was utilized ([Table 2]). T2*  relaxation times were calculated from inline-reconstructed T2*  maps by using a pixelwise, monoexponential nonnegative least-squares fit analysis (SyngoTM MapIt; Siemens Ltd.). For morphological assessment, a sagittal turbo inversion recovery measurement (TIRM) sequence and an axial proton-density-weighted fat-saturated TSE sequence (PDfs TSE) were obtained ([Table 2]) and subsequently assessed for chondral/osteochondral lesions [27] and classified using the MRI-modified Outerbridge grading system [28].

Table 1 Timeline of measurements M (t1-t4): mean (SD), range.

Measurement time points M(t1-4)

t1

t2

t3

t4

Stage [No.]

16–22

29–35

43–46

50–58

Total distance run to M [km]

1235 (104)

2127 (91)

3033 (70)

3708 (175)

1068–1487

1985–2362

2964–3161

3430–4037

Total run time to M [h]

144.2 (26.5)

251.1 (47.5)

352.2 (57.9)

438.1 (75.7)

108–185

198–358

279–470

341–624

Distance run since last M [km]*

1235 (104)

910 (132)

907 (107)

674 (169)

1068–1487

645–1170

720–1099

466–1073

Running time since last M [h]

144.2 (26.5)

118.3 (30.0)

233.9 (47.4)

204.2 (59.0)

108–185

82.6–178.8

154.6–314.4

136.7–332.6

Last stage distance before M [km]

65.9 (6.1)

67.1 (13.7)

73.5 (13.1)

77.0 (15.8)

60.1–80.5

44.0–85.0

60.3–85.7

59.5–95.1

Last stage running time before M[h]

8.0 (1.3)

7.8 (2.1)

8.3 (1.4)

9.3 (1.7)

5.6–9.5

5.1–12.7

6.0–10.3

5.7–11.5

Time elapsed between stage finish and MRI at M [hrs]:

3.3 (1.4)

2.7 (1.2)

2.5 (1.1)

2.5 (0.9)

- right hindfoot

1.2–6.1

1.2–5.0

1.5–4.5

1.0–4.2

- left hindfoot

3.5 (1.4)

2.8 (1.1)

2.7 (0.9)

2.7 (0.8)

1.4–5.8

1.5–4.7

1.8–4.7

1.3–4.0

Body mass (BM) [kg/m 2 ]

70.6 (10.7)

68.6 (10.6)

68.4 (10.2)

68.2 (9.7)

53.5–91.9

51.7–90.4

51.3–89.4

51.9–88.0

Body mass index (BMI) [kg/m 2 ]

22.6 (2.1)

21.9 (2.1)

21.9 (1.2)

21.8 (1.9)

20.6–28.4

20.1–27.9

20.3–27.6

20.0–27.2

*For NF, the time point has not been recorded because subjects stayed at a checkpoint until transportation. Therefore, their last resting period was consistently assumed to start at noon.

Table 2 MRI protocol.

Sagittal T2* GREa

Sagittal TIRM

Coronal PDfs TSE

Flip angle [°]

60

140

150

Echo time [ms]

4.5/12.2/19.9/27.7/35.4

60

32

Repetition time [ms]

1010

5830

Inversion time [ms]

120

Slice thickness [mm]

2.5

2

3

Distance factor [%]

10

20

30

Field of view [mm]

135×135

300×300

160×160

Matrix size [pixel]

320×320

512×512

384×384

Pixel size [mm], in-plane resolution

0.422 iso 0.178 mm2

0.586 iso 0.343 mm2

0.415 iso 0.172 mm2

Pixel bandwidth [Hz/pixel]

250

130

150

Total acquisition time [min:sec]

4:15

5:37

3:46

a: syngoTM MapIt; Siemens Ltd., Erlangen, Germany


#

Image postprocessing

The regions of interest (ROIs) for T2*  values were manually drawn on the sagittal T2*  GRE slices, as illustrated in [Fig. 1d]. To analyze T2*  values, six ROIs were placed in the opposing articular cartilage in each of the subtalar and midtarsal joints: a) TCJ (posterior, middle, anterior facets): talar and calcaneal; and b) TNJ: talar and navicular, CCJ: calcaneal and cuboidal medial/lateral ([Fig. 1b and c]) articular cartilage, respectively. The mean T2*  values of each ROI were utilized for further analysis. Cartilage thickness was measured in three areas of each ROI (exception: only one measurement in the medial CCJ), and the mean was taken for further analysis ([Fig. 1e]). All measurements and assessments were performed by three musculoskeletal radiologists (UHS, BC, DS; 15, 8, and 2 years of experience): UHS did the intraobserver analysis, UHS and BC did the interobserver analysis and DS did the analysis on the repeated measurement at baseline and during TEFR under supervision of UHS.

Zoom Image
Fig. 1 a: 3D scheme of (7) a: posterior facet, b: middle facet, c: anterior facet of the subtalar joint (TCJ; talocalcaneal joint) and midtarsal joints (MTJ); (8) d: talonavicular joint (TNJ) and (9) e: calcaneocuboid joint (CCJ), respectively, with MRI of b: medial and C: lateral aspects of the hindfoot in the sagittal plane; T2* w GRE, d: fused colored T2*  GRE relaxation time maps (syngo™ MapIt fusion technique), e: cartilage thickness [mm] measurements (3 in each slice), (1) tibia, (2) talus, (3) calcaneus, (4) navicular bone, (5) cuboid and (6) ankle/tibiotalar joint.

#

Statistical analyses

For statistical analysis, SPSSTM (IBMTM Statistics, release 19.0, 2010, SPSS Inc.) was utilized. Intraobserver precision was determined with two measurements at t0 three weeks apart (n=120 areas mean T2* , n=320 cartilage thickness). For inter-rater reliability, measurements of two investigators were compared (n=240) and Bland-Altman scatter plots [4] were generated with the 95% limits of agreement (LOA=precision: mean difference±1.96 SD). Correlation coefficients on rater reliabilities lambda were calculated as proposed by Jepsen et al. [18]. A t-test for independent variables was used to detect differences between opposing joint surface T2*  values. For all tests a P-value of 0.05 indicated significance. Differences between T2*  values at fixed time points were calculated by one-way ANOVA and pairwise post hoc analysis on the significance between the different ROIs: Scheffé test was used if variance homogeneity was provided, otherwise Bonferroni’s least significant difference test was applied. To analyze significant changes of T2*  and thickness values between t0 and time points (t1–t4) in the course of TEFR, a one-way ANOVA for repeated measurements was utilized. Given the longitudinal nature of the test data, a general linear model for repeated measurements was applied. Bonferroni procedure for confidence-interval adaption was applied. To determine a linear or polynomial trend of T2*  values, significance of inner-subject effects for the ANOVA were calculated. Differences between the final measurement (t4) and any maximal peak during the race (t1–t3) were compared to baseline (t0), and a paired-samples t-test with calculation of the effect size according to Cohen [7] was applied (large d=0.8, medium d=0.5, small d=0.3). Regression analysis using a linear mixed model for fixed effects was utilized to define cofactors of running burden, including body mass (BM) and body mass index (BMI).


#
#

Results

Measurement accuracy

Intra- and interrater reliability for ROI sizes, cartilage thickness, and T2*  measurements ( [Table 3] ). A: The intra- and interrater reliabilities of mean ROI sizes were high for all single ROIs and regions ranging from 0.986 to 0.998 and 0.984 to 0.990 for the TCJ and from 0.984 to 0.990 and 0.908 to 0.989 for the MTJ, respectively. The intra- and interrater precision (=LOA) at baseline was 1.8 mm2 (7.4%) and 2.2 mm2 (8.9%) for TCJ and 2.0 mm2 (11.0%) and 2.1 mm2 (11.6%) for MTJ. B: In the TCJ the intra- and interrater precision error of cartilage thickness measurements was 0.19 mm (6.9%) and 0.20 mm (6.9%). Reliabilities, ranging from 0.981 to 0.985 were high for all three regions. In the MTJ the intra- and inter-rater reliabilities were also high for all ROIs, ranging from 0.975 to 0.994 as evidenced by high intra- and inter-rater precision error with 0.19 mm (9.1% and 8.7%). C: The intra- and inter-rater reliabilities were very high for the T2*  relaxation times for all examined hindfoot joints, ranging from 0.996 to 0.998 for TCJ and from 0.996 to 0.999 for MTJ, respectively. This is also reflected in a high intra- and inter-rater precision error for corresponding cartilage regions with 0.7 ms (3.3% and 3.4%) for TCJ and 0.6 ms (3.0%) and 0.7 ms (3.4%) for MTJ, respectively, making a T2*  value progression analysis feasible.

Table 3 Cartilage ROI sizes, mean T2*  values and thickness measurements at baseline: mean (SD), range.

subtalar joint (TC)

M

rater

n

posterior

middle

anterior

talar

calcaneal

talar

calcaneal

talar

calcaneal

ROI size [px*], only right side:

t 0

(1)

20

193.2 (49.5)

171.1 (39.1)

151.2 (30.7)

129.8 (24.9)

106.9 (23.6)

91.2 (22.4)

129–335

118–227

101–213

91–181

67–148

56–138

(2)

193.1 (48.9)

170.1 (40.4)

151.2 (29.3)

129.5 (24.4)

106.7 (23.5)

91.1 (22.3)

130–333

117–282

100–209

93–183

66–145

57–136

T2*  relaxation time [ms], only right side:

t 0

(1)

20

18.2 (4.8)

20.4 (5.3)

22.5 (4.4)

23.5 (4.9)

17.6 (3.8)

21.9 (5.7)

11.9–31.2

12.0–32.7

13.8–32.1

15.4–35.6

11.9–23.8

15.3–38.2

(2)

18.2 (5.0)

20.3 (5.3)

22.3 (4.3)

23.6 (5.1)

17.5 (3.8)

21.7 (5.9)

11.5–31.8

11.8–33.0

13.8–31.8

15.2–36.3

12.1–23.6

15.2–38.9

thickness [mm], only right side:

t 0

(1)

20

3.05 (0.51)

2.94 (0.53)

2.62 (0.44)

2.0–3.9

1.7–3.9

1.9–3.3

(2)

3.00 (0.53)

2.88 (0.53)

2.57 (0.45)

2.0–3.8

1.6–3.9

1.8–3.3

midtarsal joint (MT)

M

rater

n

talonavicular joint (TN)

calcaneocuboidal joint (CC)

talar

navicular

medial

lateral

calcaneal

cuboidal

calcaneal

cuboidal

ROI size [px*], only right side:

t 0

(1)

20

109.0 (26.2)

111.0 (22.3)

74.9 (15.4)

65.7 (17.6)

131.6 (31.1)

125.8 (27.5)

63–158

81–171

44–105

39–102

71–185

70–183

(2)

108.9 (26.2)

111.1 (22.6)

75.1 (15.5)

65.5 (17.3)

132.5 (31.7)

125.6 (27.1)

62–156

80–172

43–107

38–100

72–187

70–180

T2*  relaxation time [ms], only right side:

t 0

(1)

20

16.5 (6.0)

15.6 (4.6)

25.1 (4.4)

21.7 (3.4)

21.6 (6.9)

18.1 (5.3)

10.7–33.4

10.8–27.5

17.6–32.9

16.7–29.5

14.6–37.6

12.6–30.6

(2)

16.4 (6.0)

15.8 (4.7)

25.0 (4.4)

21.5 (3.3)

21.4 (6.8)

18.3 (5.4)

10.8–34.0

10.8–27.9

17.8–32.4

17.0–29.7

14.5–36.7

12.3–31.3

thickness [mm], only right side:

t 0

(1)

20

1.90 (0.27)

2.72 (0.45)

2.18 (0.41)

1.4–2.4

2.0–4.0

1.4–2.9

(2)

1.85 (0.26)

2.68 (0.48)

2.13 (0.42)

1.4–2.3

1.9–4.0

1.4–2.9

* px: pixel (size: 0.178 mm2), MI: measurement interval, rater (1): DS, rater (2): UHS


#

Cartilage thickness

At baseline the mean cartilage thickness in the TCJ ranged from 2.6 mm (anterior facet) to 3.1 mm (posterior facet) and in the MTJ from 1.9 mm (TNJ) to 2.7 mm (CCJ). There was no significant change or relevant trend in cartilage thickness over time (t1 to t4) during the race (p-values > 0.5 for univariate ANOVA and inner-subject contrasts; Table S1 in supplemental material).


#

T2*  relaxation times

a) T2*  at baseline

At baseline the mean T2*  values (T0) for the right and left were 18.7 ms (SD 3.1) and 21.5 ms (SD 3.4) for talar and 20.5 ms (SD 3.3) and 23.7 ms (SD 4.2) for calcaneal cartilage in the MTJ, and 20.3 ms (SD 3.9) and 22.6 ms (SD 3.4) for talocalcaneal and 17.5 ms (SD 2.8) and 18.7 ms (SD 1.6) for naviculocuboidal cartilage in the MTJ, respectively.

b) T2*  changes over time (compared to baseline)

TCJ Significant T2*  changes (in %) compared to baseline occurred in all ROIs of the TCJ with the exception of the talar anterior and posterior facets of the left TCJ. However, when combined, total anterior and posterior TCJ facet differences became significant as well ([Table 4] and [5]). When all ROIs were compared to baseline, a significant mean T2*  increase of 20.9% (talar 19.9%, calcaneal 22.2%) was observed between the 1100 and 2400 km distance run ([Table 5], [Fig. 2]). The initial bilateral increase of all single averaged ROIs of the TCJ was followed by a significant decrease of T2*  values during the second half of the TEFR with a high effect size ([Table 4],[Fig. 2]): mean of about 6.0% (7.4% talar, 5.0% calcaneal) corresponding with 28.5% of initial increase (talar 37.3%, calcaneal 22.5%). However, only on the right, the single and averaged ROIs of the TCJ showed a linear and quadratic trend of the curve with a high test power ([Table 4]). On the left TCJ, a quadratic trend of the curve could be observed for total averaged joint and the talar cartilage only, whereas single ROIs had a quadratic and linear trend in the posterior and middle calcaneal ROIs. These trends had only a medium test power ([Table 4]).

Zoom Image
Fig. 2 Relative changes of T2*  relaxation times of the anterior, middle and posterior facets of the subtalar joint (TCJ) in the finisher’s group; nF=13.

Table 4 Changes in intrachondral T2*  values [ms] in the TCJ in the course of TEFR.

side

ROIs, regions

n

Mauch-ly’s test

Test on T2* -signal changes ANOVA a

Paired t-test on secondary T2* -decrease

Test on inner-subject effects

P-value

P-value

Test power

P-value

Effect size

P-value

Test power

right

talar-posterior

12

0.330

<0.001

1.000

<0.001

1.31 b

<0.001

0.995 c

talar-middle

0.306

0.013

0.833

0.001

0.89 b

(0.015)

0.744 c

talar-anterior

0.282

<0.001

0.998

0.001

0.78

0.009

0.817 c

calcaneal-posterior

0.671

<0.001

1.000

0.002

0.98 b

0.002

0.952 c

calcaneal-middle

0.806

0.001

0.957

0.003

1.09 b

0.007

0.848 c

calcaneal-anterior

0.267

<0.001

0.999

0.002

1.27 b

0.003

0.933 c

talar all

0.339

<0.001

1.000

<0.001

1.19 b

0.002

0.966 c

calcaneal all

0.074

<0.001

1.000

0.001

1.12 b

0.001

0.969 c

total TCJ

0.157

<0.001

1.000

0.001

1.18 b

0.001

0.980 c

left

talar-posterior

11

0.026

0.294a

0.262

0.022

0.8 b

0.097

0.380

talar-middle

0.127

0.006

0.889

<0.001

0.95 b

(0.035)

0.595

talar-anterior

0.658

0.081

0.604

0.001

1.45 b

0.088

0.402

calcaneal-posterior

0.246

0.001

0.960

0.002

0.82 b

0.026

0.656

calcaneal-middle

0.267

0.004

0.911

0.002

0.91 b

0.019

0.715 c

calcaneal-anterior

0.428

0.454

0.269

<0.001

0.94 b

(0.462)

0.107

talar all

0.003

0.060a

0.564

0.008

0.86 b

0.039

0.570

calcaneal all

0.031

0.025 a

0.692

0.004

0.83 b

0.066

0.461

total TCJ

0.001

0.026 a

0.676

0.001

0.88 b

0.027

0.647

ROIs: regions of interest (talar-posterior, talar-middle, talar-anterior, calcaneal-posterior, calcaneal-middle, calcaneal-anterior); Bold fonts show significance (P-values); a: “Greenhouse-Geisser” correction procedure was used; b: significant T2*  relaxation time decrease after initial increase with large effect size (Cohen`s d > 0.8) at the end of TEFR related to max. (peak) in the first part of TEFR; c: significant and quadratic and (linear) trend with high test power.

Table 5 Absolute T2*  values of subtalar joint in the course of TEFR in the finisher’s group; nF=13.

side

ROIs, regions

n

Mean (SD), 95% CI

T0

T1

T2

T3

T4

right

talar-posterior

12

16.8 (3.9), 2.1

23.5 (2.8), 1.6

22.8 (3.1), 1.7

22.9 (3.9), 2.1

21.2 (3.2), 1.7

talar-middle

21.8 (3.6), 2.0

24.7 (3.5), 2.0

25.1 (4.1), 2.2

24.9 (3.6), 2.0

23.8 (3.7), 2.0

talar-anterior

17.5 (3.9), 2.1

23.2 (4.6), 2.6

22.5 (4.5), 2.5

21.3 (4.6), 2.5

21.7 (4.7), 2.5

calcaneal-posterior

18.9 (4.8), 2.6

26.0 (2.9), 1.7

25.0 (3.1), 1.7

24.7 (3.4), 1.9

23.2 (4.9), 2.7

calcaneal-middle

22.3 (3.9), 2.1

25.8 (1.7), 0.9

26.9 (3.6), 2.0

26.8 (3.0), 1.6

25.2 (4.1), 2.2

calcaneal-anterior

20.4 (3.1), 1.7

28.3 (4.5), 2.5

26.8 (3.8), 2.1

27.1 (3.4), 1.9

25.2 (4.9), 2.7

talar all

18.7 (3.1), 1.7

23.8 (2.5), 1.4

23.5 (2.9), 1.6

23.0 (3.1), 1.7

22.2 (3.0), 1.6

calcaneal all

20.5 (3.3), 1.8

26.7 (1.8), 1.0

26.3 (2.4), 1.3

26.2 (2.8), 1.5

24.5 (4.1), 2.2

total TCJ

19.6 (3.1), 1.7

25.2 (1.6), 0.9

24.9 (1.8), 1.0

24.6 (2.3), 1.3

23.4 (3.2), 1.7

left

talar-posterior

11

23.6 (4.9), 2.7

26.0 (5.3), 3.0

26.2 (5.2), 2.9

24.9 (4.3), 2.3

23.9 (6.1), 3.3

talar-middle

21.0 (4.3), 2.4

25.0 (2.6), 1.5

25.0 (3.0), 1.7

24.7 (3.0), 1.6

24.2 (2.6), 1.4

talar-anterior

19.9 (3.3), 1.8

22.2 (2.7), 1.6

22.3 (2.7), 1.5

21.2 (2.9), 1.6

20.8 (2.6), 1.4

calcaneal-posterior

22.2 (4.6), 2.5

27.0 (2.8), 1.6

27.6 (2.8), 1.6

26.3 (3.9), 2.1

26.1 (3.3), 1.8

calcaneal-middle

22.4 (5.2), 2.8

26.2 (3.0), 1.7

27.6 (1.6), 0.9

27.0 (2.9), 1.6

26.5 (2.5), 1.3

calcaneal-anterior

26.5 (5.1), 2.7

28.4 (2.9), 1.6

28.3 (3.9), 2.2

27.6 (4.3), 2.4

29.5 (3.5), 1.9

talar all

21.5 (3.4), 1.9

24.4 (2.7), 1.5

24.5 (2.6), 1.4

23.6 (2.6), 1.4

23.0 (2.9), 1.6

calcaneal all

23.7 (4.2), 2.3

27.2 (1.7), 1.0

27.8 (1.1), 0.6

27.0 (2.2), 1.2

27.3 (1.9), 1.0

total TCJ

22.6 (3.6), 2.0

25.8 (1.9), 1.1

26.1 (1.5), 0.8

25.3 (1.9), 1.0

25.2 (1.9), 1.0

ROIs: regions of interest (talar-posterior, talar-middle, talar-anterior, calcaneal-posterior, calcaneal-middle, calcaneal-anterior)

MTJ Regarding the MTJ functional unit of both sides, ANOVA showed a significant percent increase of the T2*  relaxation time during the first half of the race compared to baseline (t0) for all single and averaged ROIs, with the exception of the right medial ROI of the CCJ and the left talonavicular joint ([Tables 6] and [7]). The overall T2*  increase of 26.3% (talar and calcaneal cartilage 34.5%, navicular and cuboidal cartilage 22.8%) was followed by a T2*  decrease in the second half of the race of 4.2% (talar and calcaneal cartilage 7.1%, navicular and cuboidal cartilage region 3.8%) compared to a 16.0% initial increase (talar and calcaneal 20.6%, navicular and cuboidal 16.7%) with high effect size for secondary T2*  decrease ([Table 7], [Fig. 3]).

Zoom Image
Fig. 3 Relative changes of T2*  relaxation times of the midtarsal joints (MTJ), the talonavicular (TNJ) and the calcaneocuboid (CCJ) joints, respectively in the finisher’s group; nF=13.

Table 6 Changes in intrachondral T2*  values [ms] in the MTJ in the course of TEFR.

side

ROIs, regions

n

Mauch-ly’s test

Test on T2* -signal changes ANOVA a

Paired t-test on secondary T2* -decrease

Test on inner-subject effects

P-value

P-value

Test power

P-value

Effect size

P-value

Test power

right

talar

12

0.412

<0.001

1.000

0.001

1.63 b

<0.001

0.999 c

navicular

0.411

<0.001

1.000

0.001

1.28 b

<0.001

1.000 c

total TNJ

0.094

<0.001

1.000

0.001

1.50 b

<0.001

1.000 c

calcaneal-lateral

0.528

<0.001

1.000

0.002

1.31 b

0.001

0.982 c

calcaneal-medial

0.968

0.218

0.427

<0.001

0.98 b

(0.447)

0.111

cuboidal-lateral

0.728

<0.001

0.998

0.002

0.76

0.007

0.852 c

cuboidal-medial

0.185

0.991

0.063

0.004

1.22 b

(0.683)

0.067

calcaneal all

0.879

<0.001

0.996

0.003

1.33 b

0.007

0.860 c

cuboidal all

0.745

0.006

0.891

0.002

0.71

0.028

0.633 c

total CCJ

0.013

<0.001 a

0.992

0.012

0.79 b

0.007

0.852 c

talocalcaneal all

0.897

<0.001

1.000

0.008

1.49 b

0.001

0.992 c

naviculocuboidal all

0.897

<0.001

0.999

0.001

0.97 b

0.001

0.977 c

total MTJ

0.137

<0.001

1.000

0.007

1.14 b

<0.001

0.995 c

left

talar

11

0.352

0.003

0.932

0.034

0.92 b

(0.011)

0.800

navicular

0.180

0.065

0.639

0.093

0.23

(0.006)

0.873 c

total TNJ

0.020

0.020 a

0.728

0.055

0.62

(0.003)

0.939 c

calcaneal-lateral

0.539

<0.001

0.995

0.001

1.96 b

<0.001

1.000 c

calcaneal-medial

0.071

<0.001

1.000

0.028

0.84 b

0.005

0.900 c

cuboidal-lateral

0.483

<0.001

0.980

0.037

0.73

(0.003)

0.930 c

cuboidal-medial

0.259

0.001

0.971

<0.001

2.14 b

0.003

0.948 c

calcaneal all

0.105

<0.001

1.000

0.009

1.47 b

<0.001

0.998 c

cuboidal all

0.201

<0.001

0.980

0.002

1.17 b

0.013

0.779 c

total CCJ

0.030

<0.001 a

0.995

0.007

1.30 b

0.001

0.985 c

talocalcaneal all

0.276

<0.001

0.999

0.024

1.12 b

0.001

0.979 c

naviculocuboidal all

0.432

0.001

0.965

0.025

0.92 b

0.048

0.532

total MTJ

0.012

<0.001 a

0.990

0.049

0.87 b

0.005

0.901 c

ROIs: regions of interest (TNJ: talar, navicular; CCJ: calcaneal lateral, calcaneal medial, cuboidal-lateral, cuboidal-medial; MTJ, fused regions: talar and calcaneal, navicular and cuboidal); Bold fonts show significance (P-values); a: “Greenhouse-Geisser” correction procedure was used; b: significant T2* -relaxation time decrease after initial increase with large effect size (Cohen`s d > 0.8) at the end of TEFR related to max. (peak) in the first part of TEFR; c: significant and quadratic and (linear) trend with high test power..

Table 7 Absolute T2*  values of midtarsal joint in the course of TEFR in the finisher’s group; nF=13: mean (SD), 95% CI.

side

ROIs, regions

n

Mean (SD), 95% CI

T0

T1

T2

T3

T4

right

talar

12

15.6 (4.0), 2.2

24.9 (3.5), 2.0

25.6 (3.1), 1.7

23.7 (5.4), 2.9

22.9 (4.2), 2.3

navicular

14.8 (2.9), 1.6

21.0 (2.8), 1.6

21.2 (1.6), 0.9

20.2 (3.8), 2.1

19.9 (2.7), 1.5

total TNJ

15.2 (3.3), 1.8

22.9 (2.8), 1.6

23.4 (1.9), 1.0

22.0 (4.1), 2.2

21.4 (3.2), 1.8

calcaneal-lateral

19.6 (5.6), 3.0

29.6 (3.3), 1.9

28.6 (3.0), 1.6

27.5 (2.9), 1.6

26.9 (4.6), 2.5

calcaneal-medial

25.7 (4.6), 2.5

29.1 (4.8), 2.7

28.9 (3.2), 1.8

27.3 (3.9), 2.1

28.1 (3.7), 2.0

cuboidal-lateral

16.5 (4.8), 2.6

22.1 (2.2), 1.3

22.3 (3.6), 2.0

21.3 (4.6), 2.5

21.6 (4.4), 2.4

cuboidal-medial

21.4 (3.6), 1.9

21.7 (2.6), 1.5

22.1 (3.1), 1.7

22.0 (3.7), 2.0

22.3 (2.0), 1.1

calcaneal all

22.7 (4.1), 2.2

29.3 (3.2), 1.8

28.8 (2.7), 1.5

27.4 (2.5), 1.4

27.5 (3.2), 1.7

cuboidal all

19.0 (3.0), 1.6

21.9 (2.2), 1.2

22.2 (3.2), 1.7

21.7 (3.8), 2.1

21.9 (2.8), 1.5

total CCJ

20.8 (3.3), 1.8

25.6 (1.9), 1.1

25.5 (2.5), 1.4

24.5 (2.1), 1.1

24.7 (2.3), 1.3

talocalcaneal all

20.3 (3.9), 2.1

27.8 (2.4), 1.3

27.7 (1.5), 0.8

26.2 (3.2), 1.7

26.0 (3.1), 1.7

naviculocuboidal all

17.6 (2.8), 1.5

21.6 (1.8), 1.0

21.9 (2.3), 1.3

21.2 (3.3), 1.8

21.3 (2.3), 1.2

total MTJ

18.9 (3.1), 1.7

24.7 (1.5), 0.8

24.8 (1.4), 0.8

23.7 (2.2), 1.2

23.6 (2.3), 1.3

left

talar

11

24.9 (4.4), 2.4

27.3 (2.7), 1.5

28.3 (4.2), 2.4

27.6 (4.3), 2.3

30.1 (3.5), 1.9

navicular

20.9 (1.5), 0.8

22.3 (3.0), 1.7

23.1 (3.5), 2.0

23.0 (3.5), 1.9

24.0 (3.1), 1.7

total TNJ

22.9 (2.7), 1.5

24.8 (2.4), 1.4

25.7 (3.0), 1.7

25.3 (3.2), 1.7

27.1 (2.8), 1.5

calcaneal-lateral

23.4 (3.5), 1.9

29.7 (3.0), 1.7

28.6 (4.0), 2.2

29.3 (2.8), 1.5

26.7 (2.4), 1.3

calcaneal-medial

19.6 (3.5), 1.9

27.5 (5.2), 2.9

26.4 (2.9), 1.6

27.0 (3.6), 2.0

25.9 (3.2), 1.8

cuboidal-lateral

19.5 (1.8), 1.0

21.9 (2.9), 1.6

24.0 (3.2), 1.8

22.6 (3.3), 1.8

23.2 (2.9), 1.6

cuboidal-medial

15.8 (2.7), 1.5

20.9 (2.8), 1.6

21.1 (3.6), 2.0

20.9 (3.6), 2.0

18.7 (2.1), 1.1

calcaneal all

21.5 (3.3), 1.8

28.6 (3.8), 2.1

27.5 (2.0), 1.1

28.2 (2.2), 1.2

26.3 (2.2), 1.2

cuboidal all

17.7 (2.0), 1.1

21.4 (2.6), 1.5

22.6 (3.0), 1.7

21.7 (3.2), 1.7

20.9 (2.3), 1.2

total CCJ

19.6 (2.5), 1.4

25.0 (2.8), 1.6

25.0 (1.8), 1.0

25.0 (2.2), 1.2

23.6 (1.9), 1.0

talocalcaneal all

22.6 (3.4), 1.9

28.1 (2.6), 1.5

27.8 (1.6), 0.9

28.0 (2.2), 1.2

27.6 (2.0), 1.1

naviculocuboidal all

18.7 (1.6), 0.9

21.7 (2.5), 1.4

22.7 (2.8), 1.6

22.2 (2.4), 1.3

22.0 (2.0), 1.1

total MTJ

20.7 (2.4), 1.3

24.9 (2.1), 1.2

25.3 (1.4), 0.8

25.1 (1.7), 0.9

24.8 (1.6), 0.9

ROIs: regions of interest (TNJ: talar, navicular; CCJ: calcaneal lateral, calcaneal medial, cuboidal-lateral, cuboidal-medial; MTJ, fused regions: talar and calcaneal, navicular and cuboidal)

TNJ The right TNJ showed a highly significant initial mean increase ([Table 7]) of 53.7% (64% talar, 42.8% navicular), and secondary decrease of T2*  values of 13.0% (17.3% talar, 8.6% navicular) with a mainly quadratic trend compared to baseline. On the left side, the TNJ showed a nonsignificant T2*  increase of about 12.0% (13.3% talar, 10.5% navicular) compared to baseline with only significance for the talar cartilage and a significant linear trend for the TNJ in total ([Table 7], [Fig. 3]).

CCJ A T2*  increase in the right CCJ measured 23.1% (29.3% calcaneal, 17.1% cuboidal) and secondary decrease of 4.4% (8.0% calcaneal, 1.4% cuboidal). The left CCJ showed a significant initial increase of about 28.0% (33.2% calcaneal, 27.7% cuboidal) and a secondary decrease of T2*  values of about 7.3% (10.6% calcaneal, 9.1% cuboidal) with a quadratic trend during TEFR, with the exception of the right medial and the left lateral cuboidal ROI ([Table 7], [Fig. 3]).

Asymmetry Highly significant left-right asymmetry of the TCJ and MTJ with statistically significant higher T2*  changes on the right were observed (p<0.001 and 0.004, respectively) (Table S2 in supplemental material).


#

Influence of cofactors of running burden on T2*  values

Significant correlation between T2*  and total distance, run time, and number of stages was found for the entire right posterior facet of the subtalar joint (TCJ) and talar aspect of the left subtalar joint (TCJ). The left TNJ showed significant correlations with total distance run and total number of stages, but the right TNJ did not (Table S3 in supplemental material). BM/BMI and T2*  values showed no significant correlations at any time point.


#

Osteochondral lesions

Only one focal navicular osteochondral lesion (Outerbridge grade 1) with associated marrow edema was identified ([Fig. 4]). Although the osteochondral lesion remained stable in size, a steady increase of T2*  over time was observed.

Zoom Image
Fig. 4 Osteochondral lesions (yellow arrow) in the hindfoot joints. Sagittal TIRM slices, through: a posterior facet of the subtalar joint (TCJ) and b calcaneocuboid joint (CCJ) of a 53-year-old male finisher: 1: baseline, 2: stage 19/1260 km run, 3:stage 32/2176 km run, 4: stage 54/3763 km run. c talonavicular joint (TNJ) of a 60-year-old male nonfinisher: 1: baseline, 2: stage 19/1260 km run, 3:stage 32/2176 km run, 4: stage 39/2666 km run. d Talonavicular joint (TNJ) of a 54-year-old male finisher: baseline, 2: stage 17/1131 km run, 3:stage 32/2176 km run, 4: stage 53/3669 km run.

#
#

Discussion

Information on hindfoot joints is limited, largely due to the lack of in vivo methods to assess the consequences of loading. Although the effects of loading on biochemical and physiological properties of the native articular cartilage occur, these changes are usually minute when measured after a single event. In our study design, we exploited the rare and unique opportunity to monitor the consequences of extreme endurance loading, accumulated over the course of an ultramarathon that covered a distance equivalent to 100 single marathons in 64 contiguous stages.

Our most important finding was that the T2*  values within the articular cartilage of the hindfoot during the ultramarathon initially became abnormal and subsequently actively recovered during the second half of the race. Cartilage matrix recovery became evident during continued extreme loading while a previous report demonstrated recovery after a long resting period of 3 month [26]. The second important finding was the observation that changes are not symmetrical bilaterally but rather asymmetric in specific patterns, acquiring a “biochemical signature” of loading over time.

In agreement with data provided by Welsch et al. [44] for the ankle joint (coefficient of variation 3.2 to 4.7%), we demonstrated excellent reproducibility for quantitative T2*  mapping of the ROIs in the subtalar and midtarsal joint: TCJ 3.3 to 3.4%, MTJ 3.0 to 3.4%.

The two major macromolecular components of the extracellular cartilage matrix, Type II collagen and proteoglycans (PGs), are responsible for its biomechanical resilience, with collagen providing elastic properties [3] [19] [31] and PGs providing viscoelastic properties [1] [2]. Water occupies most of the interfibrillar extracellular matrix, approximately 70% of which is free to move when loaded by compressive forces [37]. Quantitative T2*  measurements are most sensitive to changes in free water and collagen orientation [30]. Increasing T2*  values reflect free water dissociated from its chemical bonding to PGs [17] and is inversely related to PG content. This has been confirmed in studies on long-distance running [21] [26]. Similarly, our result of significant T2*  elevation during the first half of the race in both the subtalar (20.9%) and midtalar (26.3%) joints can be explained by loss of structural anisotropy in the collagen matrix and a concomitant increase in free intrachondral water and a decreasing PG content.

A positive effect of joint loading on chondrocyte function with increased PG and collagen synthesis was found previously [15]. Load-bearing exercises minimize the development of osteoarthritis and increase PG content and cartilage thickness in rodent models [13] [24] and dogs [22] [35]. In human knee cartilage, an increase in GAG was detected following moderate exercise [34], and it was documented in elite runners and untrained volunteers that exercise increases PG content [42]. Increasing hydrostatic pressure upregulates PG and Type II collagen mRNA expression [43] and de novo synthesis of PGs will be initiated. This upregulation with further ultramarathon running may explain the reduction of free water as indicated by decreasing T2*  values over time due to restored structural anisotropy of the collagen matrix and recovered PG content.

The talonavicular and subtalar joints are linked anatomically and functionally. Because three functions (weight acceptance, single limb support, and advancement of the limb [36]) are completed in each gait cycle, locally different contact forces [32] and energy dissipation [36] occur within the hindfoot. With initial weight acceptance and single limb support, energy dissipation has a higher effect on the distal articular surface in our data, mostly on the calcaneal aspect of posterior facet of the subtalar joint. The main functions of the transverse tarsal joint are to ‘unlock’ the midfoot into a flexible structure to accommodate variations in terrain and dissipate energy during weight acceptance and to progressively ‘lock’ the foot into a rigid structure for efficient force transfer as the cycle advances to push-off. For push-off, the talonavicular joint revealed larger T2*  changes on the navicular site, which experiences the highest energy dissipation [5]. The asymmetry found between the left and right hindfoot indicates higher loading on the dominant limb during the ultramarathon. Bilateral asymmetry in human running is known and its relationship to limb preference and risk of injury is under investigation [5]. Based on our data, future studies utilizing MRI could provide reliable insights into running style, leg preference, and biomechanical alignment.

Furthermore, MRI articular cartilage relaxation time mapping provides a reliable tool for the detection of cartilage degeneration before radiographic or morphological MRI changes of osteaoarthritis occur. Contrary to popular belief, our findings support growing evidence that endurance running [6] [23] protects normal articular cartilage. Our data provide evidence that articular cartilage is able to regenerate under running conditions and proves that articular cartilage matrix composition is designed for loading. Therefore, walking and running will keep it healthy.

Our study had several limitations. One limitation was the relatively small number of subjects as well as a possible selection bias because the participants were all healthy individuals with ultramarathon experience prior to study enrollment. Another limitation was the lack of complementary quantitative cartilage mapping sequences such as T1rho [33] or GAG-CEST, which could not be utilized due to time and study design constraints.

In conclusion, our results demonstrated that the T2*  values within the articular cartilage of the hindfoot initially became abnormal and subsequently recovered during the second half of the ultramarathon, leading to the unexpected finding that articular cartilage is capable of active functional adaptation under extreme endurance loading. This is the first time that a recovery during continued extreme loading was detected. Our findings on the subtalar joint are in keeping with the results obtained on the tibiotalar joint articular cartilage [40], further supporting the evidence that this response is not an isolated finding, but rather a universal physiological adaptive mechanism of chondrocyte function via up-regulation of de novo synthesis of proteoglycans and collagen previously known from animal research. Furthermore, these changes occur in a distinct asymmetric pattern leaving a “biochemical signature” of articular cartilage that has been acquired over time and allows in vivo insight into distinct weight bearing and joint loading. In conclusion, the normal articular cartilage of the subtalar and midtarsal joints is resilient and adaptive, leaving extreme endurance activities up to limitless human ambition.


#
#

Conflict of Interest

The authors have no conflict of interest to declare.

Acknowledgements

We would like to thank all the athletes of TEFR who took part at this project. Considering their immense physical and mental stresses they showed an extraordinary compliance on every day of the race. Contributions. All authors of this manuscript had substantial contribution to conception and design or acquisition, analysis and interpretation of data; all revised it critically for important intellectual content and did final approval of the version to be published. Role of funding source. This work is supported in part by the German Research Association (DFG: “Deutsche Forschungsgemeinschaft”), under Grants SCHU 2514/1-1 and SCHU 2514/1-2. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. No additional external funding was received for this study. Competing interests. The authors declare that they have no financial or non-financial competing interests. There are no financial or non-financial competing interests of other people or organizations influencing our interpretation of data or presentation of information.

Supplementary Material


Correspondence

Uwe Hans-Werner Schütz, PD Dr.med.
Universitatsklinikum Ulm
Klinik für Diagnostische und Interventionelle Radiologie
Albert-Einstein-Allee 29
89075 Ulm
Germany   
Phone: +49/171/7628 003   
Fax: +49/731/500 61002   


Zoom Image
Fig. 1 a: 3D scheme of (7) a: posterior facet, b: middle facet, c: anterior facet of the subtalar joint (TCJ; talocalcaneal joint) and midtarsal joints (MTJ); (8) d: talonavicular joint (TNJ) and (9) e: calcaneocuboid joint (CCJ), respectively, with MRI of b: medial and C: lateral aspects of the hindfoot in the sagittal plane; T2* w GRE, d: fused colored T2*  GRE relaxation time maps (syngo™ MapIt fusion technique), e: cartilage thickness [mm] measurements (3 in each slice), (1) tibia, (2) talus, (3) calcaneus, (4) navicular bone, (5) cuboid and (6) ankle/tibiotalar joint.
Zoom Image
Fig. 2 Relative changes of T2*  relaxation times of the anterior, middle and posterior facets of the subtalar joint (TCJ) in the finisher’s group; nF=13.
Zoom Image
Fig. 3 Relative changes of T2*  relaxation times of the midtarsal joints (MTJ), the talonavicular (TNJ) and the calcaneocuboid (CCJ) joints, respectively in the finisher’s group; nF=13.
Zoom Image
Fig. 4 Osteochondral lesions (yellow arrow) in the hindfoot joints. Sagittal TIRM slices, through: a posterior facet of the subtalar joint (TCJ) and b calcaneocuboid joint (CCJ) of a 53-year-old male finisher: 1: baseline, 2: stage 19/1260 km run, 3:stage 32/2176 km run, 4: stage 54/3763 km run. c talonavicular joint (TNJ) of a 60-year-old male nonfinisher: 1: baseline, 2: stage 19/1260 km run, 3:stage 32/2176 km run, 4: stage 39/2666 km run. d Talonavicular joint (TNJ) of a 54-year-old male finisher: baseline, 2: stage 17/1131 km run, 3:stage 32/2176 km run, 4: stage 53/3669 km run.