Assessment of Meniscal Healing Status by Magnetic Resonance Imaging T2 Mapping After Meniscal Repair

Abstract

Background: Although the sensitivity and specificity of magnetic resonance imaging (MRI) for the diagnosis of primary meniscal tears are high, these values are lower for the assessment of healing status of repaired menisci. Purpose: To compare the accuracy of MRI T2 mapping and conventional MRI in assessing meniscal healing after repair. Study Design: Cohort study (diagnosis); Level of evidence, 2. Methods: Patients who underwent meniscal repair with concurrent anterior cruciate ligament reconstruction between 2012 and 2016 and had a follow-up second-look arthroscopy were enrolled. The patients were divided into healed and incompletely/not healed groups based on the second-look arthroscopy findings. For the repaired menisci, the following were compared between the groups, (1) Stoller and Crues classification on conventional MRI with a proton density-weighted fat-saturated sequence and (2) the remaining colored meniscal tear line on T2 mapping coincident with the high signal line showing the primary tear on conventional MRI were compared. The change of T2 relaxation time (ΔT2) of the colored meniscal tear line pre- to postoperatively was compared between the groups. The mean T2 relaxation time of the whole area of the postoperative meniscus at each slice was also compared with that of control menisci to assess the whole quality of the repaired meniscus. Results: A total of 26 menisci from 24 knees were assessed (16 healed menisci, 10 incompletely/not healed menisci). According to the Crues classification on conventional MRI, 8 of 16 healed menisci and 3 of 10 incompletely/not healed menisci improved from grade 3 to 2, with there being no significant difference between the groups (P = .43). However, the colored meniscal tear line remained in only 3 of the 16 healed menisci as compared with 9 of the 10 incompletely/not healed menisci, and the presence of this colored line allowed differentiation between healed menisci and incompletely/not healed menisci (sensitivity, 81.3%; specificity, 90.0%; odds ratio, 39.0; P = .001). The mean (SD) ΔT2 was −31.1 ± 3.2 and −19.9 ± 4.4 ms in the healed and incompletely/not healed groups, respectively (P < .001). Receiver operating characteristic curve analysis showed a cutoff ΔT2 value of −22.3 ms for separation of meniscal healing (P < .001). The T2 relaxation times of the whole area of the repaired menisci were 31.7 ± 3.4 and 32.8 ± 3.8 ms in the healed and incompletely/unhealed groups, respectively (P = .69), with these values being significantly longer than the 26.9 ± 2.2 ms in the controls (P < .001). Conclusion: MRI T2 mapping allowed the differentiation of healing status after meniscal repair, with high sensitivity and specificity as compared with conventional MRI.