Table 1.
Description of studies analyzing medial collateral ligament injury with magnetic resonance imaging (MRI)
| References | Patient knees/controls | Age (mean, range) | Medial joint space opening | MRI findings/main findings | Intraobserver/interobserver reliability | Sensitivity (%)/specificity (%) | PPV/NPV |
|---|---|---|---|---|---|---|---|
| Garvin et al. 1993 (RS) [41] | 23/– | n.r., 16–53 | n.r |
Match between surgery findings and MRI: Partial rupture (low signal intensity interrupted by areas of high signal intensity): 13 of 7 cases identified Complete rupture (low signal intensity interrupted across entire width): 8 of 14 cases identified |
n.r | 100/n.r.c | n.r |
| Yao et al. 1994 (RS) [35] | 41/22 | 31, 16–85 | n.r |
Match between clinical examinations and MRI Normal MCL: 18 of 22 cases, grade I: 7 of 13 cases, grade II: 9 of 18 cases, grade III: 7 of 10 cases |
n.r | n.r | n.r |
| Mirowitz et al. 1994 (PS) [31] | 64/10 | 35, 14–69 | n.r |
Match between clinical examinations and MRI: Grade I: 31 of 46 cases, grade II: 8 of 14 cases, grade III: 1 of 4 cases |
n.r | 81–100/40–95d |
0.25–0.63/ 0.93–1.00d |
| Rasenberg et al. 1995 (PS) [32] | 21/21 | 18, 17–50 | n.r |
Match between instrumented clinical examinations and MRI: Grade I: 13 of 14 cases, grade II: 5 of 5 cases, grade III: 1 of 2 cases |
n.r | n.r | n.r |
| Schweitzer et al. 1995 (PS) [33] | 76/25 | 26, 18–60 | n.r |
Correlation between MRI grade and clinical examination: Grade I (subcutaneous oedema): pain 44%, tenderness 56%, swelling 56%, instability 33% Grade II (morphologic disruption and/or internal high signal intensity and/or fluid in the MCL bursa): pain 80%, tenderness 87%, swelling 59%, instability 26% grade III (MCL discontinuity): pain 64%, tenderness 64%, swelling 50%, instability 18% |
n.r./0.76–0.93 |
7–81/95–100e (different signs) |
n.r |
| Lundberg et al. 1996 (PS) [40] | 69/– | 26, 13–57 | n.r |
match between arthroscopy and MRI: 14 of 69 cases true positive, 41 of 69 cases true negative, 3 of 69 cases false positive, 11 of 69 cases false negative |
n.r | 56/93b | 0.82/0.79b |
| De Maeseneer et al. 2001 (CS) [50] | 7/6 |
n.r., 50–91 Control group n.r., 16–48 |
n.r |
bursa could not be detected via MRI in cadaveric knees without contrast injection On anatomic section MCL bursa was observed in five of seven cases in the femoral and in seven of seven cases in the tibial component |
n.r | n.r | n.r |
| Wen et al. 2007 (RS) [34] | 6/12 | 47, n.r | n.r |
Traumatic MCL-oedema: oedema deep and superficial: in six of six cases, oedema fibers: in four of six cases, bone marrow oedema: in three of six cases, medial meniscal tears: in three of six cases, medial meniscal extrusion: in one of six cases Atraumatic MCL-oedema: oedema deep and superficial: in 12 of 12 cases, oedema fibers: in 5 of 12 cases, marginal osteophytes: in 6 of 12 cases, articular cartilage thinning: in 6 of 12 cases, bone marrow oedema: in 5 of 12 cases, subchondral cysts: in 1 of 12 cases, medial meniscal tears: in 11 of 12 cases, medial meniscal extrusion: in 6 of 12 cases |
n.r | n.r | n.r |
| Halinen et al. 2009 (PS) [39] | 44/– | 39, 21–64 | n.r |
Match between MRI and surgical treated MCL (n = 21): Identification of tear grade: 18 of 21 cases, identification of tear location: 11 of 21 cases |
n.r | 86a/n.r | n.r |
| Studler et al. 2011 (PS) [30] | 10/10 | 35, 17–51 |
MCL injury: grade I: 2.3 mm, grade II: 2.9 mm (1.9–4.9 mm) control group: 1.7 mm (0.7–3.3 mm) |
Grade I (n = 1): edema around an intact MCL Grade II (n = 9): partial tear of the ligament with internal high signal |
n.r./ ICC: 0.89–0.94 |
n.r | n.r |
| Taketomi et al. 2014 (RS) [47] | 12/– | 25, 16–40 | n.r |
Wave sign: in all cases Identification of the ruptured end: in 9 of 12 cases Identification of entrapment: in 2 of 2 cases |
n.r | n.r | n.r |
| Alaia et al. 2019 (RS) [48] | 65/– | Only subgroups reported | n.r |
Distal tibial grade III sMCL tear: in 20 of 65 cases Isolated tibial attachment tears: in 16 of 20 cases Femoral and tibial attachment tears: in 4 of 20 cases SLL: in 12 of 20 cases Borderline SLL: 6 of 20 cases |
n.r | n.r | n.r |
| Brimmo et al. 2019 (PS) [45] | 7/– | 24, 16–32 | n.r |
SLL: redundant distal MCL fibers, displaced superficial to the pes anserinus Clinical examination: grade II in 1 of 7 cases, grade III in 6 of 7 cases |
n.r | n.r | n.r |
| Boutin et al. 2020 (RS) [49] | 51/– | 28, n.r | n.r |
SLL: in 20 of 51 sMCL tears Wave sign: in 18 of 20 SLL-cases and in 21 of 31 non-SLL-cases Proximal sMCL stump is located more distal and medial in cases with SLL |
n.r | n.r | n.r |
CS cadaveric study, FS fat-suppressed, FSE fast spin echo, FSFS fat-suppressed fluid-sensitive, ICC intra-class correlation coefficient, MCL medial collateral ligament, NPV negative predictive value, n.r not reported, PD proton density, PPV positive predictive value, PS prospective study, RS retrospective study, SLL Stener-like-lesion, sMCL superficial medial collateral ligament, T1w T1-weighted MRI, T2w T2-weighted MRI, 0.5 T 0.5 Tesla MRI, 1.5 T 1.5 Tesla MRI, 2D FLASH two-dimensional fast low-angle shot, 3 T 3 Tesla MRI
aArthroscopy findings (no further definition)
bArthroscopy findings: increased opening of the medial compartment
cSurgery findings
d, eClinical examination