A Prospective Study of the Nonoperative Treatment of Degenerative Meniscus TearsBy Todd Rimington, MD; Krishna Mallik, MD; Douglas Evans, MD; Kenneth Mroczek, MD; Bruce Reider, MD
ORTHOPEDICS 2009; 32:558
This prospective study was designed to evaluate a nonoperative-based treatment protocol for patients with the clinical diagnosis of a degenerative medial meniscus tear and to determine if a significant percent of patients with degenerative medial meniscus tears improve with nonoperative treatment and do not elect operative treatment. Twenty-six patients were followed for 37 months and evaluated clinically, radiographically, and with standardized, validated Modified Lysholm Knee Scoring System (MLKS) and Standardized Activities of Daily Living Scale of the Knee (SADL) questionnaires. Patients were initially treated with 4 weeks of nonsteroidal anti-inflammatory drugs. After 4 weeks, they were offered arthroscopic partial meniscectomy or continued nonoperative treatment. Forty-six percent of patients (12 of 26) declined operative treatment. The mean length of time between enrollment and surgery was 3 months (range, 1-13 months). Both groups improved significantly over baseline (P<.05). The initial and final SADL and initial MLKS scores of the 2 groups were not significantly different (P>.05). The final MLKS score of the operative treatment group was significantly greater than the nonoperative group (P=.04). Both the nonoperative and operative treatment groups improved significantly at 3-year follow-up. Forty-six percent of the patients received nonoperative treatment and improved to a functional level that allowed them to avoid surgery. Therefore, we recommend an initial course of nonoperative treatment for all patients with a clinical diagnosis of degenerative medial meniscus tears prior to considering surgery.
Meniscus tears represent one of the most common injuries to the knee and may be the most common indication for knee surgery in the adult patient. The incidence of meniscus injuries treated by meniscectomy has been estimated at 61 per 100,000.1 Meniscus injuries can be divided into 2 groups: traumatic tears and degenerative tears. Traumatic tears occur in a younger population and are usually the result of a discrete traumatic episode. Degenerative tears are thought to progress from intrasubstance degeneration within the menisci.1 Magnetic resonance imaging (MRI) has shown intrasubstance degeneration of the menisci in patients as young as 30 years.2
The diagnosis of meniscus injury continues to evolve and is based on clinical and radiographic findings. Multiple tests have been suggested to enhance the diagnosis of meniscus injury by physical examination, including posterior joint line tenderness, McMurray’s test, and the Apley compression and distraction test, among others.3-6 In studies comparing these tests, joint line tenderness has been 60% to 90% sensitive and specific.4,6-8 Magnetic resonance imaging diagnosis of degenerative meniscus tears has also been shown to have a high degree of accuracy.9-11 Previous studies have shown similar efficacy in diagnosis between clinical examination and MRI.7-12 Other studies conclude MRI is most useful in patients with a history consistent for a meniscus tear but an inconclusive physical examination.12,13
Numerous studies have evaluated the effectiveness of operative treatment of meniscus tears14-18; however, there is a paucity of literature evaluating the effectiveness of nonoperative treatment. To date, we have found only 1 prospective study of nonoperative treatment of meniscus tears.19 Therefore, we designed a prospective study to evaluate a nonoperative-based treatment protocol for patients with the clinical diagnosis of a degenerative medial meniscus tear and to determine if a significant percentage of patients with degenerative medial meniscus tears improve with nonoperative treatment and do not elect operative treatment. We hypothesized that a clinically significant percentage of patients will improve to a satisfactory level with nonoperative treatment and will decline operative treatment.
Materials and Methods
After receiving Institutional Review Board (IRB) approval, all patients in the senior author’s (B.R.) practice who met the inclusion criteria were offered the option to enroll in the study. Inclusion criteria included: age 35 to 60 years, no history of significant acute trauma, no previous surgery on the affected knee, medial joint line tenderness with no abnormal ligament laxity on physical examination, normal weight-bearing radiographs, and the ability to take nonsteroidal anti-inflammatory drugs (NSAIDs). An IRB-approved written consent was obtained from all patients wishing to participate in the study.
Study participants were evaluated clinically and radiographically. Radiographs were performed to rule out osteoarthritis, including a weight-bearing anteroposterior view of both knees, a weight-bearing 30° flexed-knee posteroanterior view of both knees, a lateral view, and a 30° axial view of the affected knee. Participants were followed with 2 previously validated knee questionnaires: the Modified Lysholm Knee Scoring System (MLKS)20 and the Standardized Activities of Daily Living Scale of the Knee (SADL).21 The MLKS is a knee-specific questionnaire evaluating the pain, function, and swelling of the knee. The MLKS questionnaire has 19 questions with a high score of 100. The SADL is a knee-specific scale evaluating the way in which the state of a patient’s knee affects his or her daily function. The SADL questionnaire has 17 questions with a total of 80 points possible and is converted to a score out of 100.
All patients enrolled in the study had a closed MRI (1.5 Tesla) performed on the knee. Patients who did not present to the initial clinic visit with an MRI had an MRI performed before their second clinic visit. A standard musculoskeletal knee MRI protocol was used, consisting of sagittal proton density T2-weighted spin-echo images, coronal T1-weighted images, coronal T2-weighted images with fat saturation, axial proton density with fat saturation, and sagittal proton density with fat saturation images. All MRIs were read by a fellowship-trained musculoskeletal radiologist at our institution, including MRIs performed at outside institutions.
At the initial appointment, the patient was given a prescription for naproxen 500 mg twice daily until the next appointment 4 weeks later. If the patient had a history of gastrointestinal sensitivity, rofecoxib 25 mg daily was used. At 4-week follow-up, the questionnaire was re-administered. If the patient’s symptoms improved, treatment was discontinued, and the patient was followed up at 1 month and then re-evaluated at 3 months. Improvement was determined by both subjective assessment and objective improvement on the MLKS and SADL questionnaires.
If the patient did not improve satisfactorily after 1 month of NSAIDs, arthroscopy was offered. If the patient chose to continue nonoperative treatment, the NSAIDs were continued and follow-up was scheduled for 1 month. At each of the follow-up appointments, if symptoms persisted, arthroscopy was offered.
For patients who chose surgery, a routine arthroscopy of the knee was performed using 3 portals (anteromedial, anterolateral, and superolateral) by the senior author (B.R.) under regional block anesthesia. All encountered pathology was treated at the time of the arthroscopy. If a meniscus tear was found, partial meniscectomy was performed; the meniscus tear was debrided, conserving as much stable meniscus remnant as possible. The findings at surgery were documented with regard to meniscus injury as well as other intra-articular findings. Patients were allowed to bear weight as tolerated on the involved extremity postoperatively. Crutches were used for stability postoperatively and were discontinued by the patient when he or she could ambulate independently.
Whether patients chose surgery, all were followed prospectively with MLKS and SADL scores approximately every 3 months via mailings. Nonoperative patients could opt for surgery at any time during the follow-up period.
For the purpose of data analysis, patients were divided into 2 treatment groups: a nonoperative treatment group and an operative treatment group. Each patient’s SADL and MLKS scores were recorded; special attention was given to the initial and final scores for both scales. The initial scores represented the status of the patients’ knees before any treatment was given. The final scores represented the status of the patients’ knees after treatment was completed. The two-sample equal variance, dependent Student t test was used to evaluate differences within treatment groups. Significance was determined at P<.05. The assumptions for the Student t test were satisfied, as the results of the SADL and MLKS were shown to be normally distributed with the Anderson-Darling and Shapiro-Wilk tests of normality with α<0.05, and the groups tested equally variant with the F test with α<0.05.
From April 1999 to April 2005, 32 patients were enrolled in the study. Six patients were lost to follow-up before completing 4 weeks of nonoperative treatment; therefore, they were excluded from the final data analysis. Twenty-six patients were included in the final data analysis. Eighteen men and 8 women had a mean age of 47.5 years (range, 36-57 years). Eighty-eight percent of patients (23 of 26) were treated with naproxen and 12% (3 of 26) were treated with rofecoxib. Mean follow-up was 37 months (range, 10-59 months). Twelve patients (12 knees) were managed with nonoperative treatment and 14 patients (15 knees) opted for operative intervention. There were 15 right knees and 12 left knees. One operative patient had bilateral medial meniscus tears and opted to have bilateral meniscectomies. The SADL and MLKS scores from this patient were treated as a single patient in the questionnaire analysis. Therefore, 46% of patients (12 of 26) were treated nonoperatively and 54% (14 of 26) were treated operatively.
The nonoperative group had 12 patients: 5 men and 7 women. The operative group had 14 patients: 13 men and 1 woman. Mean age for the nonoperative group was 48.2 years (range, 41-56) and for the operative group was 47.0 years (range, 36-57 years). The nonoperative group was followed for a mean of 38 months (range, 10-67 months), and the operative group was followed for a mean of 37 months (range, 10-59 months). There was no significant difference in age (P=.65) and length of follow-up (P=.92) between the 2 groups. Mean time between enrollment and surgery was 3 months (range, 1-13 months).
The initial and final SADL scores for all patients were 66.2 and 87.2, respectively. The initial and final MLKS scores for all patients were 59.7 and 85.2, respectively. A significant improvement of P<.001 was observed between the mean initial and final SADL and MLKS scores of all patients. Individually, 77% of patients (20 of 26) showed overall improvement on both the SADL and MLKS scores, 8% (2 of 26) showed improvement on 1 scale and decline on the other scale, 8% (2 of 26) showed improvement on 1 scale and remained the same on the other scale, and 8% (2 of 26) showed a decline on both scales. Of the patients who opted for nonoperative treatment, 75% (9 of 12) showed improvement on both scales, whereas 79% of patients (11 of 14) who opted for operative treatment improved on both scales.
The initial and final SADL and MLKS scores for the nonoperative group and the operative group are shown in Table 1. There was no significant difference in the mean initial and final SADL and the initial MLKS scores between the 2 groups (P<.05; Table 2). The mean final MLKS score of the operative group was significantly greater than the nonoperative group (P=.04; Table 2).
Magnetic resonance imaging was performed on 27 knees in 26 patients during the study (Table 1). Magnetic resonance imaging diagnosed a meniscus tear in 81% of knees (22 of 27). In the patients who chose operative treatment, the MRI diagnosis was accurate in 86.7% of knees (13 of 15), and clinical diagnosis of a meniscus tear was accurate in 93.3% (14 of 15). One operative patient was found to have an intact medial meniscus but chondral injuries to the femoral trochlea, medial femoral condyle, and patella during arthroscopy; MRI had incorrectly diagnosed a medial meniscus tear for this patient. Five patients showed no medial meniscus tear on MRI. Interestingly, 4 of these 5 patients showed abnormalities within the meniscus: 1 showed degenerative changes of the medial meniscus, 1 degenerative changes of the medial meniscus and a lateral meniscus tear, 1 an irregular medial meniscus, and 1 a popliteal cyst. Four of these 5 patients were satisfied with nonoperative treatment, including the patient with the popliteal cyst. The patient who elected operative treatment had an MRI diagnosis of an irregular medial meniscus (Figure 1) and was found to have a medial meniscus tear during arthroscopy (Figure 2).
Degenerative medial meniscus tears represent a large portion of knee injuries that require surgery in the middle-aged patient.14,16,17,22 This study prospectively evaluated patients with a clinical diagnosis of a degenerative medial meniscus tear and their response to a treatment protocol that included an initial 1-month course of NSAIDs. Patients with degenerative medial meniscus tears typically present with joint line pain and an atraumatic history. Osteoarthritis can present with similar symptoms and is a common cause of pain in the knee in this patient population. Patients with both osteoarthritis and meniscus tears have poorer outcomes when treated with arthroscopic meniscectomy and may in some cases be better suited for high tibial osteotomy or total knee arthroplasty.16,22 This study focused on isolated degenerative meniscus tears; thus it was important to rule out osteoarthritis in the patients’ knees. Weight-bearing radiographs of the knee are highly effective at diagnosing osteoarthritis.23,24 A series of 4 weight-bearing knee radiographs were used to rule out osteoarthritis in the patients.
The clinical diagnosis of medial meniscus tears is highly accurate.4,6-8 Despite this, it has become routine for some orthopedic surgeons to confirm the diagnosis of a meniscus tear with MRI. Magnetic resonance imaging has been found to be highly sensitive and specific for detecting meniscus tears9-11; however, in this study and previous studies, clinical examination was as effective as MRI at diagnosing a meniscus tear.7,12 The goal of this study was to evaluate a nonoperative treatment protocol based on a clinical diagnosis of a degenerative meniscus tear. For this reason, the inclusion criteria of the study did not include an MRI finding of a meniscus tear, and patients were not excluded from the final analysis based on MRI findings. In fact, if the 4 patients treated nonoperatively with MRIs negative for meniscus tears and the single patient who had no meniscus tear at arthroscopy were excluded from the final analysis, it would not significantly changed the results. Both groups still significantly improved overall from initial to final SADL and MLKS scores (P<.05), and 40% of patients elected nonoperative treatment. Therefore, based on the results of the study, the nonoperative treatment of degenerative meniscus tears can be started based on clinical diagnosis alone.
The surgical treatment for degenerative meniscus tears, arthroscopic partial meniscectomy, has been shown to have good to excellent results in the majority of patients.16,18 However, as this study shows, a significant portion of patients with degenerative meniscus tears improve to satisfactory levels with nonoperative treatment and do not proceed with operative treatment. We are aware of only 1 other study comparing operative vs nonoperative treatment of meniscus tears.19 In that study, patients with degenerative medial meniscus tears were randomized to arthroscopic partial meniscectomy or to a nonoperative exercise program and followed prospectively. At 6-month follow-up, both groups improved significantly from baseline, but no significant difference was seen between the 2 groups. The patients in our study also improved significantly with both nonoperative and operative treatment, and the operative group had a significantly higher final MLKS score than the nonoperative group.
Interestingly, the operative treatment group in our study had a larger proportion of men; the significance of this finding is difficult to assess. It suggests that men are more likely to elect operative treatment than women. It is possible that degenerative meniscus tears in men are less amenable to conservative treatment or that men are simply less likely to be satisfied with the results of conservative treatment. Future studies are needed to better characterize which patients will benefit from operative treatment.
In this study, mean time to operative treatment for the operative group was 3 months. Given that the overall follow-up of our study was 38 months, this suggests that patients who elect operative treatment will do so with the first few months of nonoperative treatment. This is helpful in determining appropriate follow-up intervals for patients with degenerative meniscus tears. It suggests that more frequent monthly follow-up should occur initially, and if patients elect to continue nonoperative treatment, less frequent follow-up is appropriate. Until we can accurately predict who will fail nonoperative treatment, an initial course of nonoperative treatment should be recommended for every patient suspected of having a degenerative medial meniscus tear.
The strengths of this study include the simplicity of the treatment protocol and the prospective design. The treatment protocol can be easily incorporated into an orthopedic surgeon’s practice. Joint line tenderness was chosen as the basis of the clinical examination in these patients because it has been shown to be highly effective in the past and is easily reproducible.4,6-8 Although previous studies have shown that combining joint line tenderness with other clinical tests for meniscus pathology enhances the sensitivity and specificity for meniscus tears,3-6 this study’s results show that when joint line tenderness is present in patients meeting the inclusion criteria, it is highly effective at diagnosing a degenerative meniscus tear. This study’s average follow-up was relatively short at 3 years; however, it was significantly longer than the 6-month follow-up of the only other prospective study of nonoperative treatment of meniscus tears in the literature.19
Limitations of this study include small sample size, large number of patients lost to follow-up, and its non-randomized design. The small number of patients enrolled in the study and the large number of patients lost to follow-up limited the power of the results. There was a significant difference between the final MLKS score of the operative group over the nonoperative group, and although the final SADL score of the operative group was higher than that of the nonoperative group, it was not significant. With a larger sample size, this may have proven to be a significant difference between the 2 groups. The study’s lack of randomization introduced selection bias, as the patients were able to self-select operative treatment; therefore, results from the comparisons between the 2 treatment groups should be considered with caution due to the presence of selection bias.
This study provides prospective evidence of the effectiveness of a nonoperative-based treatment protocol for degenerative medial meniscus tears. Both the nonoperative and operative treatment groups improved significantly at 3-year follow-up. In addition, the operative group showed a significantly greater improvement than the nonoperative group on the MLKS. More importantly, 46% of patients received nonoperative treatment and improved to a functional level that allowed them to avoid surgery. Patients who elect operative treatment fail nonoperative treatment within the first few months; therefore, we recommend an initial 1-month course of nonoperative treatment for all patients with a clinical diagnosis of a degenerative medial meniscus tear prior to considering surgery.
Dr Rimington is from the Department of Orthopedics, Georgetown University, Washington, District of Columbia; Dr Mallik is from the Department of Orthopedic Surgery, Medical College of Ohio, Toledo, Ohio; Dr Evans is from the Department of Orthopedic Surgery, Loyola University Chicago, and Dr Reider is from the Department of Orthopedic Surgery, University of Chicago, Illinois; and Dr Mroczek is from New York University Orthopaedic Surgery Associates, New York, New York.
Drs Rimington, Mallik, Evans, Mroczek, and Reider have no relevant financial relationships to disclose. Financial support for the study was received from the University of Chicago Department of Orthopedics research fund.
Correspondence should be addressed to: Todd Rimington, MD, Georgetown University Department of Orthopedics, 3800 Reservoir Rd, Ground PHC, Washington, DC 20007.
James M. Fox, MD