T1ρ relaxation time of the meniscus and its relationship with T1ρ of adjacent cartilage in knees with acute ACL injuries at 3 T

Summary

Objective

The aim of this study was to assess the reproducibility of meniscus T1ρ measurements, and to study T1ρ relaxation time in the lateral meniscus (LM) and its relationship with adjacent cartilage T1ρ in knees with acute anterior cruciate ligament (ACL) injuries at 3 T magnetic resonance imaging.

Method

Quantitative assessment of the meniscus and cartilage was performed in 15 healthy controls and 16 ACL-injured patients using a T1ρ mapping technique. All ACL-injured patients were imaged prior to surgery within 1–3 months of injury. The anterior and posterior horns of LM and medial meniscus (MM) were associated with partitioned weight-bearing cartilage sub-compartments (anterior, central, and posterior).

Results

T1ρ measurements in the meniscus showed excellent reproducibility (coefficient of variation (CV) < 5%). Significantly elevated T1ρ values were found in the LM in patients compared with controls (P < 0.01). No differences were found in the MM. Significantly higher T1ρ values were found at the posterior horn compared with the anterior horn of patients’ meniscus (P = 0.005). At the posterior sub-compartment of lateral tibia (LT), significantly increased cartilage T1ρ values were found in patients compared with controls (P = 0.002). A significant correlation (R² = 0.47, P = 0.007) was found between T1ρ values of posterior horn of LM and T1ρ values of posterior sub-compartment of LT cartilage in patients.

Conclusion

A strong injury-related relationship was demonstrated between meniscus and cartilage biochemical changes. T1ρ mapping techniques provide tools to quantitatively evaluate meniscus and cartilage matrix in patients with ACL injuries.

Introduction

Anterior cruciate ligament (ACL) injuries are one of the most common ligament injuries of the knee joint. ACL tears are frequently associated with the damage of other structures within the knee such as the meniscus, articular cartilage, and subchondral bone. Recent long-term studies have demonstrated that 50–70% of the ACL-injured patients have radiological changes of osteoarthritis (OA) at 10–15 years after ACL injury, despite ACL reconstruction^1,2. Both orthopedic and radiology literature have reported a significant association between the meniscus tears and ACL injury, and that these ACL injuries are more commonly associated with the lateral meniscus (LM) tears than medial meniscus (MM) tears³.

The meniscal fibrocartilage structure is very similar to that of hyaline cartilage, both containing mainly water, collagen, and proteoglycans (PG). However, the hyaline cartilage has a PG concentration of 5–10%, whereas in the meniscus there is a lower PG concentration of 1–2%, as previously reported⁴. In OA, the articular cartilage and meniscal biochemical changes are preceded by the damage to the collagen–PG matrix⁵.

Magnetic resonance imaging (MRI) techniques have the potential to reflect changes in biochemical composition of the hyaline cartilage. Specifically, T1ρ mapping^6–8, and delayed gadolinium-enhanced MRI of cartilage (dGEM-RIC)^9,10 techniques can be used to study the PG content and distribution in knee joint, with the attention focused lately on early detection of OA.

To date, little is known about the effect of meniscal injury on the risk of disease progression in OA patients. Studies of the meniscectomy suggest the importance of the meniscal function loss as a risk factor for subsequent knee OA¹¹. MRI has been widely used as a sensitive and specific tool in detecting, evaluating and monitoring the meniscal injury in patients with ACL tears^12,13. Using semi-quantitative MRI with arthroscopic correlation, Hunter et al.⁵ demonstrated a strong association of the meniscal pathology changes with cartilage loss in symptomatic knee OA. In a more recent study¹⁴, using the dGEMRIC technique, significant correlations between T1(Gd) of the meniscus and T1(Gd) of the articular cartilage were found, potentially demonstrating associated degenerative processes in the knee joint.

The aim of this retrospective cross-sectional study was to (1) quantify the T1ρ relaxation time in the meniscus, and to assess its reproducibility, and (2) to study the relationship between the T1ρ relaxation time of the meniscus and adjacent articular cartilage in ACL-injured patients and healthy controls. We hypothesized that T1ρ values are significantly elevated within the injured meniscus compared with the non-injured ones, and T1ρ elevations are correlated between cartilage and meniscus in acutely ACL-injured knee joints.

Materials and methods

SUBJECTS

Two groups of subjects were recruited for this study: 15 healthy controls – four females, 11 males, age range 19–57 years (average = 30.1 ± 8.6 years) and body mass index (BMI) of 23.9 ± 2.2 – without any clinical symptoms of OA or other knee injuries, and 16 ACL-injured patients – five females, 11 males, age range 20–56 years (average = 32.5 ± 5.8 years) and BMI of 23.5 ± 1.9. The controls were physically active subjects, recruited in order to match the demographic data of the ACL-injured patients (age, BMI). All patients were imaged prior to surgical reconstruction within 3–12 weeks of injury (average = 7 weeks), and prior to ACL reconstruction. The meniscal damage was assessed using a modified semi-quantitative, multi-feature scoring method: Whole-ORgan Magnetic resonance imaging Score (WORMS)¹⁵. The anterior and posterior horns of each of the LM and MM were graded from 0 to 4 based on the sagittal T2-weighted fast spin-echo (FSE) fat-saturated images, with 0 = intact menisci, 1 = signal abnormality, 2 = non-displaced tear or prior surgical repair, 3 = displaced tear or partial resection, and 4 = complete maceration/destruction or complete resection. The study was performed in accordance with the rules and regulations the Committee for Human Research at our institution. Informed consent was obtained from all the subjects after the nature of the examinations had been fully explained.

MRI PROTOCOL

MRI of the study knee of each subject was acquired using a 3 T GE Excite Signa MR scanner (General Electric, Milwaukee, WI, USA) with a transmit/receive quadrature knee coil (Clinical MR Solutions, Brookfield, WI, USA).

Morphologic imaging

The imaging protocol included sagittal 2D T2-weighted fat-saturated FSE images [repetition time (TR)/echo time (TE)= 4300/51 ms, field of view (FOV) = 14 cm, matrix = 512 × 256 slice thickness = 2.5 mm, gap = 0.5 mm, echo train length = 9, bandwidth (BW) = 31.25 kHz, number of excitations (NEX) = 2] and sagittal three-dimensional (3D) water excitation high-resolution spoiled gradient-echo (SPGR) imaging (TR/TE = 15/6.7 ms, flip angle = 12, FOV = 14 cm, matrix = 512 × 512, slice thickness = 1 mm, BW = 31.25 kHz, NEX = 0.75). The T2-weighted fat-saturated FSE images were used to visualize the clinical aspects of the ACL injury (meniscus, and cartilage signal and morphology), and the high-resolution SPGR images were used for meniscus and cartilage segmentation.

T1ρ relaxation time mapping

Sagittal 3D T1ρ-weighted images were acquired based on spin-lock techniques and 3D SPGR acquisition¹⁶. The sagittal 3D T1ρ-weighted imaging sequence was composed of two parts: magnetization preparation for imparting T1ρ contrast, and an elliptical-centered segmented 3D SPGR acquisition immediately after T1ρ preparation during transient signal evolution. The duration of the spin-lock pulse was defined as time of spin-lock (TSL), and the strength of the spin-lock pulse was defined as spin-lock frequency (FSL). The number of pulses after each T1ρ magnetization preparation was defined as views per segment (VPS). There was a relatively long delay (time of recovery, T_rec) between each magnetization preparation to allow enough and equal recovery of the magnetization before each T1ρ preparation. The main parameters of this sequence were as follows: FOV = 14 cm, matrix = 256 × 192, slice thickness = 3 mm, TR/TE = 9.3/3.7 ms, BW = 31.25 kHz, VPS = 48, T_rec = 1.5 s, TSL = 0/10/40/80 ms, FSL = 500 Hz, total acquisition time approximately 13 min.

MR images post-processing

All images were transferred to a Sun Workstation (Sun Microsystems, Mountain View, CA, USA) for off-line data processing. Semi-automatic meniscus and cartilage segmentation was performed on the sagittal SPGR images using in-house software¹⁷ developed with Matlab (Mathworks, Natick, MA, USA) based on Bezier splines and edge detection. T1ρ maps were then reconstructed using a Levenberg–Marquardt mono-exponential in-house developed fitting algorithm. T1ρ-weighted images intensities obtained for different TSL were fitted pixel-by-pixel to the following equation: S(TSL) ∝ exp(−TSL/T1ρ). Next, the reconstructed T1ρ maps were rigidly registered to the previously acquired high-resolution T1-weighted SPGR images using the VTK CISG Registration Toolkit¹⁸.

The weight-bearing lateral femoral condyle (LFC) and medial femoral condyle (MFC) portions as well as lateral tibia (LT) and medial tibia (MT) compartments were further divided into three regional sub-compartments, similar to a previous method described by Peterfy et al.¹⁵: anterior, central and posterior [Fig. 2(A)] The anterior sub-compartment corresponds to the region above (in femur) or under (in tibia) the anterior horn of the meniscus, the sub-compartment central corresponds to the centrally uncovered region between the anterior and posterior horns of the meniscus, and the posterior sub-compartment corresponds to the region above (in femur) or under (in tibia) the posterior horn of the meniscus.

Fig. 2.

(A) Meniscus (blue) and cartilage segmented sub-compartments – anterior (green), central (red) and posterior (yellow) – displayed on a SPGR image. (B) T1ρ color map overlaid on a fat-saturated T2-weighted FSE image obtained in a patient, with higher T1ρ values; a meniscal tear (arrow) is evident at the posterior horn of meniscus.

Statistical analysis

Mean and standard deviation SD of T1ρ values were calculated in each of the meniscus and cartilage compartments in all subjects. The CVs characterizing the reproducibility of meniscus T1ρ measurements were assessed in four control subjects based on two repeated scans, based on previously reported paper¹⁹. Paired t tests were employed to compare the mean intra-group T1ρ relaxation time values within all defined sub-compartments. Standard t tests were used to compare the mean inter-group T1ρ relaxation time values within all defined sub-compartments and to compare the T1ρ relaxation time values in ACL-injured patients with and without meniscal injuries. Spearman’s rank correlations were performed between the meniscus and cartilage T1ρ values.

Results

T1ρ MEASUREMENTS REPRODUCIBILITY IN THE MENISCUS

The computed CVs showed good measurement precision for T1ρ in meniscus: 4.6% for the anterior horn and 3.3% for the posterior horn at the lateral side, and 3.7% for the anterior horn and 4.9% for the posterior horn at the medial side.

CONTROL SUBJECTS DATA

In the meniscus, no significant difference was found between T1ρ values of the anterior (15.44 ± 4.36 ms) and posterior (14.69 ± 2.36 ms) horns of the LM as well as between the anterior (14.49 ± 2.26 ms) and posterior (15.21 ± 3.14 ms) horns of the MM. No significant correlation was found between T1ρ of the meniscus and cartilage in controls.

In the articular cartilage, a significant increase in T1ρ values for the healthy controls was noticed in the LFC sub-compartments from LFC-anterior (32.03 ± 4.89 ms) to LFC-central (37.59 ± 2.77 ms) and to LFC-posterior (40.10 ± 3.00 ms). At the LT, significant differences were found between all the sub-compartments; however, the T1ρ values at LT-central (31.77 ± 2.79 ms) were lower compared to the LT-anterior (34.67 ± 4.13 ms) and the LT-posterior (37.55 ± 3.14 ms). For the medial side of the knee, no significant differences between the sub-compartments were noted at the MFC and the MT.

ACL-INJURED PATIENTS DATA

Meniscus assessment

There were no meniscal tears at the anterior horn of both LM and MM, as assessed using the WORMS scoring¹⁵. At the posterior horn, there were six meniscal tears of grade 1 and four of grade 2 in the LM, whereas seven meniscal tears of grade 1 and only one of grade 2 were found in the MM of the 16 ACL-injured patients (Table I). Three out of 12 patients required LM debridement, whereas only one patient required meniscus repair using all-inside technique. The remaining LM tears were either stable or healed at the time of surgery. Bone marrow edema-like lesions (BMEL) were present at the lateral side of the knee joints within all the ACL-injured patients. Among them, 13 patients (81%) had BMEL located at the LT, nine patients (56%) had BMEL located at the LFC, and seven patients (44%) had BMEL both at the LT and LFC. Fig. 1 shows an illustration of an ACL tear associated with meniscal tear and BMEL at the lateral side of the knee joint.

Table I.

Meniscal tears assessment of a modified semi-quantitative, multi-feature scoring method: WORMS; within this patient cohort, the meniscal tears were only present at the posterior horn of both LM and MM

Meniscal tears
		Grade 1	Grade 2
LM	Anterior horn	0	0
	Posterior horn	6	4
MM	Anterior horn	0	0
	Posterior horn	7	1

Fig. 1.

(A) Sagittal T2-weighted fat-saturated FSE images revealing an ACL tear (arrow). (B) A meniscal tear at the posterior horn of LM (arrow) as well as BMEL are observed (asteriks).

At the LM, significantly elevated T1ρ values were found in the posterior horn (21.13 ± 5.00 ms) compared with the anterior horn (17.49 ± 3.01 ms). At the MM, there was no significant difference between the anterior horn (15.31 ± 2.96 ms) and the posterior horn (15.97 ± 3.65 ms).

A more detailed meniscus assessment was performed for comparing the T1ρ data in ACL-injured patients with and without meniscal tears (Table II). Significantly elevated T1ρ values were found at the posterior horn of the LM (P = 0.0004) in patients with meniscal tears compared with patients without meniscal tears. No significant difference was found at the MM (P = 0.121).

Table II.

T1ρ (ms) data for meniscus in ACL-injured patients: comparison between patients with (+) and without (−) meniscal tears. Significant differences between with (+) and without (−) meniscal tears were only found at the posterior horn of the LM (P = 0.0004), but not in the MM (P = 0.121)

T1ρ (ms) in Posterior horn
LM		MM
+	−	+	−
23.5	17.7	17.7	14.9
3.6	4.1	4.1	2.4
P = 0.0004		P = 0.121

Articular cartilage assessment

Similarly, as in controls, a significant increase in T1ρ values was observed at the LFC sub-compartments from the LFC-anterior (31.00 ± 2.76 ms) to the LFC-central (37.25 ± 3.14 ms) and the LFC-posterior (40.54 ± 3.62 ms). At the LT, significantly increased T1ρ values were found in the LT-posterior (44.18 ± 6.81 ms) compared to the other two sub-compartments – LT-anterior (34.62 ± 7.85 ms, P < 0.001) and LT-central (33.75 ± 7.31 ms, P < 0.001). For the medial side of the knee, no significant difference was observed between all sub-compartments.

In order to perform a more detailed data analysis, T1ρ data from patients with and without meniscal tears were compared, as shown in Table III. Significantly increased T1ρ values were only found at the posterior sub-compartment of the LT (P = 0.019), and the anterior sub-compartment of the MFC (P = 0.015).

Table III.

T1ρ data for cartilage in ACL-injured patients: comparison between patients with (+) and without (−) meniscal tears. Significant differences between with (+) and without (−) meniscal tears were only found for articular cartilage at the posterior compartment of LT (P = 0.019), and the anterior compartment of MFC (P = 0.015), marked in the table (data marked in gray)

Anterior		Central		Posterior		Anterior		Central		Posterior
+	−	+	−	+	−	+	−	+	−	+	−
LFC						LT
30.7	31.4	37.1	37.6	40.8	40.1	36.0	32.3	34.3	32.2	45.9	40.1
2.8	2.7	3.5	2.5	4.4	2.1	9.2	4.5	9.1	2.3	6.3	1.8
MFC						MT
43.3	33.3	41.6	34.7	40.8	35.8	34.9	33.6	34.1	34.3	35.3	34.2
8.5	5.5	11.3	4.9	8.7	5.1	4.2	5.7	8.1	5.9	7.3	4.7

To illustrate the T1ρ data within the meniscus and cartilage regions of interest, a T1ρ color-coded map obtained in an ACL-injured patient is shown in Fig. 2(B) displaying variations of the meniscus and cartilage structure at the lateral side of the knee.

RELATIONSHIP BETWEEN T1ρ RELAXATION TIME IN ACL-INJURED PATIENTS AND CONTROL SUBJECTS

The comparison between patients and controls is shown in Fig. 3 for the lateral side, and in Fig. 4 for the medial side of the knee. For the articular cartilage, no significant difference was found between patients and controls within the femoral condyles’ sub-compartments as well as for the MT. T1ρ values were significantly higher at the LT-posterior sub-compartment for the patients compared to the healthy controls (44.18 ± 6.81 ms vs 37.55 ± 3.14 ms, P = 0.002).

Fig. 3.

T1ρ data from the lateral side of the knee: comparison between ACL-injured patients and controls, both for meniscus and cartilage.

Fig. 4.

T1ρ data from medial side of the knee: comparison between ACL-injured patients and controls, both for meniscus and cartilage.

In the meniscus, significantly increased T1ρ values were found in the patients compared with controls for both anterior (P = 0.004) and posterior (P < 0.001) horns of the LM. No significant difference between the patients and controls was found at the MM.

In patients, a significant correlation (R² = 0.47, P = 0.007) was found at the lateral side of the knee between T1ρ of the posterior horn of the meniscus and T1ρ of the posterior sub-compartment of the tibial cartilage (Fig. 5).

Fig. 5.

A significant correlation was found at the lateral side of the knee between T1ρ values of the posterior horn of the meniscus and T1ρ values of posterior sub-compartment of lateral tibial cartilage, demonstrating a strong injury-related relationship between the two tissues.

Discussion

In this study, a T1ρ mapping technique was employed to characterize the matrix of the meniscus and adjacent articular cartilage in ACL-injured knees within 1–3 months of the injury. An example of an ACL tear is illustrated in Fig. 1(A) obtained in a patient using the fat-saturated T2-weighted FSE imaging sequence. Meniscal tears were present in 12 ACL-injured patients (75%) on both sides of the knee, whereas BMEL were identified at the lateral side only in all 16 patients. These clinical results support other studies of the acute ACL injuries that have reported these lesions as predominantly located on the lateral side of the joint^20–22. In addition to these observations, a recent study²³ emphasized the consistent presence of the cartilage damage of the posterior lateral tibial plateau as well as to posterior horn tears in LM, when acute ACL injury is shown.

A significant variation in cartilage T1ρ values was evident in the three sub-compartments of the lateral side of the normal and injured knees (Fig. 3), while the medial side exhibited a homogenous distribution of the T1ρ values (Fig. 4). However, in patients, the T1ρ values for the cartilage were, in general, slightly higher compared to controls, indicating a lower PG content and/or a higher water content, as previously reported²⁴. These results suggest that early changes within the cartilage structure were initiated at the time of injury.

As previously reported^6,25, the T1ρ distribution in articular cartilage is inversely correlated with the PG content. A study that focused on the extraction of PG from human meniscus indicated a lower total glycosaminoglycan (GAG) concentration, and therefore a lower PG content in the meniscus than in articular cartilage⁴. In addition, significant regional variations in GAG content (higher in the inner zones and lower in the outer regions) have been demonstrated in porcine and bovine menisci²⁶. Taking this into account, i.e., with a lower PG content in the meniscus than in cartilage, a higher T1ρ value is expected in the meniscus compared to the articular cartilage. Our results, however, showed lower T1ρ values in the meniscus when compared with those of the articular cartilage. The biophysical basis of the T1ρ in meniscus thus is different. It is known that one of the major differences between meniscus fibrocartilage and hyaline cartilage is the predominance of type I collagen in menisci compared with the cartilage which is composed primarily of type II collagen²⁷. Also, experimental models with enzymatically degraded cartilage have shown, beside the strong correlation of the T1ρ with PG content, some minimal dependency of the T1ρ on collagen content²⁸, and water content²⁴, though some experiments have noted an influence from interactions due to the collagen fibril orientation²⁹. In another study focused on articular cartilage⁶ it was shown that there might be some effect of the collagen fibril orientation which could affect the T1ρ values computation; however, this orientation effect is being found reduced and almost absent in T1ρ due to spin-locking. Regarding the meniscus, it is not clear if there is an orientation effect and how this effect might influence the T1ρ quantification, and this specific point requires further investigations. We speculate that, besides the influence of PG content, there might be an additional contribution of the collagen (content, fiber orientation), and hydration to the T1ρ characteristics in different tissues.

The average CV of T1ρ measurements of the meniscus was less than 5%, showing a good reproducibility of the technique. However, the study is limited with only four subjects with two repeated measurements. No trend in the T1ρ values for the menisci was observed in healthy controls, both at the lateral and medial sides. Analyzing the ACL injury location – lateral side of knee joint – a significant elevation in T1ρ values was found at the posterior horn compared with the anterior horn. For a more detailed analysis, data from ACL-injured patients with and without meniscal tear were assessed. As the meniscal tears were only found at the posterior horn of both LM and MM, the analysis was done only for these particular locations. As expected, increased T1ρ values were found for the patients with meniscal tears at the posterior horn of the LT, in accordance with the acute injury location – the lateral side of the knee joint. In addition, data for the articular cartilage shown also increased T1ρ values at the posterior sub-compartment of the LT, and interestingly, at the opposite location, the anterior sub-compartment of the MFC for the patients with meniscal tears compared with patients without meniscal tears. Moreover, the T1ρ elevation from the posterior horn of the LM significantly correlated with the cartilage T1ρ elevation at the posterior sub-compartment of the LT (P = 0.007). This aspect reflected the degeneration within the hyaline cartilage matrix, and also a strong injury-related relationship between the meniscus and cartilage biochemical changes, as previously reported^3,5. Nishimori et al.²³ reporting clinical findings based on MRI and arthroscopy in patients with acute ACL injuries emphasized the importance that needs to be attributed to the cartilage damage of the posterior lateral tibial plateau as well as to the posterior horn tears in the LM. Krishnan et al.¹⁴, using the dGEMRIC technique, also found significant correlations between the T1(Gd) of the meniscus and T1(Gd) of the articular cartilage at both lateral and medial sides of OA knee, suggesting that the corresponding degradative processes occurred in both tissues. Also, based on fluorescence detection, Handl et al.³⁰ observed a higher concentration of the advanced glycation end products due to inflammatory and/or degenerative processes in the cartilage; in addition, the acute pathological changes due to diseases such as meniscal lesions or ACL rupture caused a significant increase of formation of the advanced glycation end products even in the group of young patients. The authors stated that, such an observation could be crucial and important for the detection of knee conditions suspected of early meniscal and/or ACL lesions especially among young patients.

The natural history of the ACL-deficient knee is not yet fully understood, and few prospective studies exist that assess the long-term outcome of the condition^22,31,32. Acute ACL-injured patients are predisposed to develop radiographic and symptomatic OA; therefore identifying the risk factors may help to develop preventive interventions targeted toward these patients. In this cross-sectional study we emphasized the important role of the meniscus in patients with ACL tears, since the findings might suggest that abnormal meniscal function could have potential consequences for cartilage damage. These patients will be followed longitudinally to evaluate any progressive meniscal changes and the relationship with cartilage changes and OA development.

In conclusion, T1ρ mapping technique was applied to evaluate the meniscus and articular cartilage composition in patients with ACL injuries. This noninvasive method represents a potential means of quantitatively examining the time course of meniscal tissue change in knee OA development and progression. However, the biophysical basis for T1ρ assessment relative to the meniscus molecular structure needs further investigations. A better understanding of the interaction of molecular changes in the meniscus and adjacent cartilage can lead to a critical evaluation of the current treatment and subsequently improve the outcomes following ligament injuries to the knee.