Effectiveness of Rehabilitation for Knee Osteoarthritis Associated With Isolated Meniscus Injury: A Scoping Review

Meniscus tear is the most common type of injury to the meniscus and occurs more frequently on the medial compartments than the lateral compartments. Further, it is often caused by trauma or degenerative processes and can occur anywhere on either the meniscus, anterior horn, posterior horn, or midbody. Treatment of meniscus injuries is likely to greatly impact the evolution of osteoarthritis (OA) as meniscus injuries can gradually progress to knee OA. Hence, treatment of these injuries is important for managing the progression of OA. While the types of meniscus injuries and symptoms have been reported previously, the effectiveness of rehabilitation according to the degree of meniscus injury (e.g., vertical, longitudinal, radial, and posterior horn tears) remains unknown. In this review, we aimed to investigate whether rehabilitation for knee OA associated with isolated meniscus injuries varies with the degree of injury and determine the effects of rehabilitation on outcomes. We searched PubMed, Cumulative Index to Nursing and Allied Health Literature, Web of Science, and Physiotherapy Evidence Database for studies published before September 2021. Studies on ≥40-year-old patients with knee OA and isolated meniscus injury were included for analysis. The types of meniscus injury were classified as longitudinal, radial, transverse, flap, combined, or avulsion of the anterior and posterior roots of the medial meniscus, and assigned knee arthropathy grades of 0-4 according to the Kellgren-Lawrence classification. The exclusion criteria were meniscus injury, combined meniscus and ligament injury, and knee OA associated with combined injury in patients <40 years of age. There were no restrictions on the region, race, or gender of participants, or language or research format of the studies. The outcome measures were the Knee Osteoarthritis Outcome Score, Western Ontario and McMaster Universities Osteoarthritis Index Score, Visual Analog Scale or Numeric Rating Scale, Western Ontario Meniscal Evaluation Tool, International Knee Documentation Committee Score, Lysholm Score, 36-Item Short-Form Health Survey, one-leg hop test, timed up and go test, and re-injury and muscle strength. A total of 16 reports met these criteria. In studies that did not classify or distinguish degrees of meniscus injury, the effects of rehabilitation were generally favorable in the medium-to-long term. In cases where the intervention was not sufficiently effective, patients were recommended either arthroscopic partial meniscectomy or total knee replacement. Studies on medial meniscus posterior root tear did not confirm the effectiveness of rehabilitation due to the short intervention period. Further, Knee Osteoarthritis Outcome Score cut-offs, clinically important differences in Western Ontario and McMaster Universities Osteoarthritis Index, and minimum important changes in patient-specific functional scales were reported. Of the 16 studies reported in this review, nine met the definition. This scoping review has a few limitations such as the effect of rehabilitation alone could not be examined, and the intervention effectiveness differed at short-term follow-up. In conclusion, there was a gap in evidence regarding the rehabilitation of knee OA after isolated meniscus injury due to differences in intervention duration and methods. In addition, on short-term follow-up, intervention effects varied across studies.

Research has not yet indicated whether meniscus degeneration is a causative factor for knee OA. However, knee arthritis begins with deterioration of the meniscus involving degenerative lesions and progresses to OA of the knee due to loss of meniscus function. Moreover, OA may cause extrusion of the meniscus and degenerative lesions in the knee joint as well as accelerating structural progression [8]. Morphological deformities of the meniscus (extrusion) and meniscus incompleteness (tears) are strongly related to the incidence and progression of knee OA [9]. Treatment of meniscus injuries is likely to play an important role in managing the progression of OA as meniscus injuries can eventually lead to knee OA. Meniscus tears and knee OA are known to cause pain and other symptoms. Treatment involves either conservative or surgical intervention [10,11]. Treatment of knee OA associated with meniscus injuries is initially conservative. Surgical treatment is preferred as the next option if the conservative treatment proves to be ineffective [5]. Rehabilitation, a form of conservative therapy, reportedly reduces pain and restores physical function in patients with knee OA [12,13]. Rehabilitation for medial and lateral meniscus injuries has been reported to be effective in the short and long term [14][15][16]. However, the difference in the effects of rehabilitation according to the degree of damage to the medial and lateral menisci (e.g., vertical, longitudinal, radial, and posterior horn tears) remains unclear.
While some studies have reported on the types of meniscus injuries and symptoms, the effectiveness of rehabilitation according to the degree of meniscus injury remains unknown. Therefore, this scoping review was conducted with two objectives. The primary objective was to assess whether knee OA with and without isolated meniscus injuries exhibits different rehabilitation outcomes according to the extent of medial and lateral meniscus injuries. The secondary objective was to determine how short-term and long-term outcomes after treatment change over time based on the extent of meniscus injury.

Overview
This scoping review was conducted to identify research findings related to isolated meniscus injury and knee OA and was conducted according to The Joanna Briggs Institute scoping review methodology [17] and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Extension for Scoping Reviews (PRISMA-ScR) guidelines [18]. This scoping review defined participants, concepts, and context (PCC) as requirements. The requirement for ethical approval of this study was waived considering the lack of participant involvement in this review. The review protocol was submitted to the "protocols.io" database for publication (dx.doi.org/10.17504/protocols.io.6qpvrd5w3gmk/v2).

Eligibility Criteria: Participants, Concepts, and Context Criteria
Types of participants: Patients aged ≥40 years with knee OA with isolated unilateral or bilateral meniscus injuries were included irrespective of their sex [2]. The inclusion criteria were as follows: knee OA grades 0-4 as defined by the Kellgren-Lawrence (KL) classification and traumatic or degenerative isolated meniscus injuries; the types of meniscus injuries were longitudinal, radial, horizontal, flap, and compound tears, as well as avulsion of the medial meniscus anterior and posterior root tears. The exclusion criteria were as follows: meniscus injuries in individuals aged <40 years, combined meniscus and ligament injuries (anterior cruciate ligament, posterior cruciate ligament, medial collateral ligament, and lateral collateral ligament), or knee OA associated with combined injuries. Additionally, cases with locking and catching due to meniscus injury, cartilage loss associated with traumatic meniscus injuries, patellofemoral OA, and surgical treatment (meniscectomy and repair) or orthotic therapy (immobilization in the acute phase of meniscus injury) were also excluded.
Different study designs were targeted to identify gaps in evidence. Specifically, we included interventional studies (cluster randomized controlled trials (RCTs) and RCTs), observational studies (cohort, crosssectional, and longitudinal studies), and case reports. Systematic reviews, meta-analyses, and narrative reviews were excluded.

Search Strategy
A comprehensive electronic search for studies on meniscus injuries was conducted using PubMed, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Web of Science, and Physiotherapy Evidence Database (PEDro) databases. Further, sources of gray literature were searched using Open Gray [19]. A complete search strategy for these five databases was developed using keywords from the titles and abstracts of the relevant articles ( Table 1). To conduct a comprehensive search for meniscus injuries, the search strategy was developed without including keywords related to OA. For a comprehensive literature search, we selected studies published from the inception of each database to September 20, 2021. OR MH "tibial meniscus injuries") AND (MH "physical therapy modalities" OR TI "physical therapy" OR AB "physical therapy" OR TI physiotherapy OR AB physiotherapy OR TI kinesiotherapy OR AB kinesiotherapy OR MH rehabilitation OR TI rehabilitation OR AB rehabilitation OR MH "resistance training" OR TI "resistance training" OR AB "resistance training" OR TI "strength training" OR AB "strength training" OR TI "neuromuscular training" OR AB "neuromuscular training" OR MH "exercise therapy" OR TI "exercise therap*" OR AB "exercise therap*" OR TI "exercise program*" OR AB "exercise program*" OR TI "exercise training" OR AB "exercise training" OR TI "aerobic training" OR AB "aerobic training" OR TI "aerobic exercis*" OR AB "aerobic exercis*" OR TI "training program*" OR AB "training program*" OR TI "resistive exercis*" OR AB "resistive exercis*" OR TI "resistive training" OR AB "resistive training" OR MH "endurance training" OR TI "endurance training" OR AB "endurance training" OR TI "endurance exercis*" OR AB "endurance exercis*" OR TI Instructio* OR AB Instructio*) Web of Science search strategy (meniscus OR "tibial meniscus injuries") AND ("physical therapy modalities" OR "physical therapy" OR physiotherapy OR kinesiotherapy OR rehabilitation OR "resistance training" OR "strength training" OR "neuromuscular training" OR "exercise therapy" OR "exercise program*" OR "exercise training" OR "aerobic training" OR "training program*" OR "resistive exercis*" OR "resistive training" OR "endurance training" OR "endurance exercis*" OR Instructio*)   [20]. Screening was assessed against the inclusion criteria for titles and abstracts after pilot testing by two independent reviewers (MH and SK). Additionally, selected full-text articles were evaluated against comprehensive criteria by two independent reviewers (MH and SK). Any disagreements between the reviewers at any stage of the selection process were discussed and resolved by the authors (MH and SK). If an agreement could not be reached, a third reviewer (TK) was consulted to resolve the issue.

Data Extraction
Data were extracted from the selected papers using Microsoft Excel® for Microsoft 365. Data extraction was specific to PCC as well as to study design findings relevant to the purpose of this scoping review. Data extracted included the author's name and year of publication, country where the study was conducted, study design, study population (age and sex), sample size, purpose of the study, type of meniscus injury (medial and lateral meniscus, medial meniscus anterior, and posterior root tear), knee OA (KL classification), type and duration of intervention (including follow-up duration of intervention), outcomes, adverse events, and study limitations [21]. Additionally, for studies that compared rehabilitation with surgical treatment, only rehabilitation was considered for data extraction. Data extraction was conducted by two reviewers (MH and SK), and disagreements were discussed and resolved between them. In case of further disagreement, a third reviewer (TK) was consulted to adjudicate on the issue. As necessary, the authors of the original papers were notified if the information required for peer review was missing or if additional data were requested.

Data Analysis and Integration of Results
The PRISMA flowchart was used to present the search results and process of incorporation in a graphical form [22]. Additionally, a table summarizing the study characteristics, intervention and follow-up periods, intervention effects, adverse events, and limitations is presented. An online tool was used to create the diagram (https://www.mapchart.net/).

Meniscus Tear or Symptoms
The rehabilitation programs in the studies that did not classify or distinguish the degree of meniscus injuries included warming up, neuromuscular, and balance exercises around the knee and hip joints, as well as muscle strengthening exercises (Table 3).
Author  The duration of exercise varied widely with each study, ranging from five to 12 weeks. The follow-up periods also varied considerably, from one month to five years (    There were no restrictions due to participation in leisure-time physical activities during the study period, and no differences were reported in terms of type, frequency, or intensity of leisuretime physical activities between the groups. Leisure-time physical activities were self-reported at follow-up, which is limited by recall bias and overestimating fitness level. There is a large difference in time between baseline and intervention initiation between the two groups.   Additionally, both serious and non-serious adverse events following rehabilitation were reported. Serious adverse events included neurological problems, cardiovascular and other systemic conditions, reoperation on the affected knee, and TKR [13,24,28,35]. In contrast, non-serious adverse events that were reported included PT and exercise-induced falls, knee joint swelling, and lower extremity pain [13,24,28,35]. Importantly, limitations were reported in the studies included in this scoping review; many reported difficulty in blinding the examiner, a common limitation for rehabilitation interventions [13,28,35]. Additionally, several studies reported small sample sizes and short follow-up periods [33,36].

Medial Degenerative Meniscus Injuries
Rehabilitation for medial degenerative meniscus injuries was mainly programmed with neuromuscular and strength exercises (Table 3) [16,23,31]. The average duration of intervention by exercise was 12 weeks, with a relatively wide range of follow-up periods from three months to five years. Outcomes indicating the effects of exercise were pain and symptoms (KOOS subscale pain), physical function (muscle strength [isokinetic knee peak torque], OLH test, and 6-minute timed hop test), and ADL or QOL (KOOS subscale ADL) ( Table 4) [16,23,31]. Intervention tended to improve short-and long-term pain; other KOOS symptoms such as swelling, restricted range of motion, and mechanical symptoms; physical function; and ADL or QOL. Specifically, symptoms such as swelling, restricted range of motion, and mechanical symptoms [37].

Medial Meniscus Posterior Root Tear
Rehabilitation using MMPRT was mainly programmed with neuromuscular, strength, and ROM exercises ( Table 3) [25][26][27]29,32]. In addition to rehabilitation, medications administered included celecoxib and other NSAIDs, paracetamol, and tramadol [26,27,32]. Outcomes indicating the effects of exercise were pain and symptoms (VAS), physical function (Lysholm Knee Score), and ADL or QOL (IKDC score, KOOS subscales [sport/recreation, ADL and QOL], Tegner activity scale, and Lysholm activity scale). Some reports demonstrated significant short-term differences in intervention effects in pain and symptoms, physical function, and ADL or QOL, while others did not exhibit any differences. Specifically, Neogi et al. observed a significant difference of 0 mm (0-40, p = 0.03) in the final follow-up VAS test and 10 mm (0-70, p = 0.04) in the final activity follow-up [32]. In contrast, Ikuta et al. discovered that the VAS was -27.0 mm (-42.0 to -11.9, p = 0.004) for adapting alignment exercise, while muscle training and exercise had a symptom subscale KOOS of 15.7 points (6.4-24.9, p = 0.001); the results differed according to outcomes [29].  [25]. The limitations of these studies were small sample sizes and no medium-to long-term follow-up [25,29,32]. In contrast, a study that used combined rehabilitation and pharmacotherapy reported that it did not examine the effects of pharmacotherapy [32].

Cut-Off Scores and Clinically Important Differences
Three studies reported a clinically important difference (CID) in outcomes of meniscus injury and knee OA. Kise [28]. This review compared the data of the included studies using above the cut-off, CID, and MIC thresholds, which are listed in Table 5. Because the results of pain and symptoms, physical function, ADL, sport/recreation, and QOL are presented in this scoping review, the relevant items were summarized. In summary, two reports [16,30] evaluated the KOOS subscale for pain, one

Discussion
This scoping review summarized the effects of rehabilitation according to the degree of meniscus injury and described the existing gaps in the literature. We also summarized the effects of rehabilitation, focusing on pain, physical function, and ADL or QOL. We aimed to describe how these outcomes are influenced by rehabilitation in the short term, medium term, and long term. Studies that did not classify or distinguish the degree of meniscus injury have reported that the effects of rehabilitation were generally favorable in the medium-to-long-term duration. In contrast, for cases in which the intervention was not sufficiently effective, patients reportedly subsequently underwent APM or TKR. Additionally, studies on MMPRT have not provided a definite conclusion on the efficacy of rehabilitation due to the short intervention period. Conversely, cut-off values for assessment scores, CID, and MIC were reported only in studies that did not classify or differentiate the degree of meniscus injury. This review focused on these gaps in evidence and describes the potential areas that need to be addressed in the future.

Gaps in Research
Differences in interventions: There were differences in intervention procedures and methods among the exercises in the study that did not distinguish between the degree of meniscus injury and MMPRT exercises.
Most of the exercise programs in the interventions that did not differentiate the degree of meniscus injury initiated with a warm-up involving a stationary bike and included exercises to improve physical function, focusing on strength, neuromuscular exercises, aerobic conditioning, functional mobility, and balance exercises [13,24,28,30,[33][34][35][36]. Stationary biking was also programmed as a cool-down exercise at the end of each session [13,24,28]. In contrast, rehabilitation programs for MMPRT focused primarily on ROM exercises, muscle strength, endurance, and flexibility [25][26][27]29,32]. In addition to exercise, medication was also prescribed [26,27,32]. We consider this a gap because of the difference in methods between interventions that do not distinguish the degree of meniscus injury and MMPRT. Medications administered for MMPRT are considered an important intervention because they are recommended by the Osteoarthritis Research Society International guidelines [11]. However, to prove the effectiveness of rehabilitation alone, other interventions, such as pharmacotherapy, should be excluded. We believe that rehabilitation should be a stepwise program that improves physical function and reduces pain and symptoms [35,38,39].

Differences in outcomes:
In studies that did not distinguish the extent of meniscus injury, pain/symptoms (EQ-VAS, KOOS subscales [pain and symptoms], and VAS or NRS), physical function (IKDC score, muscle strength, PSFS, ROM, and WOMAC physical-function score), and ADL or QOL (the EQ-5D, KOOS subscales [ADL, sport/recreation, and QOL], and SF-36 physical-activity scores) were utilized [13,24,28,30,[33][34][35][36]. However, for MMPRT, the IKDC score, KOOS, and VAS, plus the Lysholm knee score and TAS score, were included as outcome measures. The Lysholm knee and TAS scores are used to evaluate outcomes of knee ligament surgery (e.g., anterior cruciate ligament, posterior cruciate ligament, medial collateral ligament, and lateral collateral ligament) and meniscal repair/meniscectomy [39][40][41]. Hence, we believe that the KOOS and WOMAC scores should be considered in lieu of the Lysholm knee or TAS scores as they can be used to determine the overall effects of the interventions on pain, physical function, ADL, and QOL. Although the overall intervention effects could not be evaluated in this scoping review, the cut-off, CID, and MIC were important clinical outcome measures in knee OA associated with meniscus injuries, which were discussed in five studies [16,29,[33][34][35]. Additionally, apart from the cut-off, CID, and MIC, some studies on knee OA reported minimal CID (MCID) in the NRS, SF-36, VAS, and WOMAC scores [42]. Specifically, they reported an NRS of 1.0, an SF-36 (physical function) score of 3.3 points, a VAS of -8.4 to -19.9 mm, and a WOMAC (pain/physical function) score of -9.7/-9.3 points [42]. Moreover, minimal clinically important improvement (MCII) was reported as -19.9 mm for VAS and -9.1 (−26.0%) points for WOMAC physical function [43]. Conversely, some studies involving middle-aged and older patients with meniscus injuries have reported IKDC scores with MIC defined as 10.9 points [44]. Hence, we examined the studies reported in this review that met these definitions, and nine papers were included in this category [13,26,27,29,[32][33][34][35][36]. The MCII is reportedly unaffected by age, duration of disease, or sex, and we believe that this is a useful definition that can be utilized in clinical practice [45]. Based on the above, we believe that the outcomes of knee OA associated with meniscus injuries should be evaluated by the KOOS subscales (e.g., pain/symptoms, physical function, ADL, sport/recreation, and QOL) or WOMAC scores, with the cut-off values, CID, MIC, MCID, or MCII applied in clinical practice.
Differences in follow-up periods: Medium-to long-term follow-up from one to five years demonstrated generally favorable changes over time [13,16,24,[26][27][28]34,35]. In contrast, when the follow-up period was shorter (between three and six months), studies reported that outcomes differed depending on the rehabilitation modality [28]. Additionally, previous studies have reported that long-term non-operative therapy may, in some cases, delay total knee arthroplasty (TKA) [46,47]. Furthermore, it was reported that patients with knee OA who avoided surgery for five years after the onset of symptoms might have a worse prognosis than those who underwent TKA. Avoidance of surgery is not necessarily an indicator of the success of non-operative treatment in these patients [46]. In contrast, two-thirds of patients could delay TKR surgery for at least two years following non-surgical treatment for moderate-to-severe knee OA [48]. There is a difference in efficacy outcomes between short-and long-term follow-ups. When the follow-up was shortterm, no consistent results were observed in the effects of the intervention. For long-term follow-ups, the response to intervention was generally positive, but a number of patients might transition to TKA/TKR. Furthermore, long-term non-operative management can delay TKA/TKR. Considering the above, when determining the length of follow-up, a comprehensive decision should be made based on the patient's symptoms, needs, and outcome indicators to determine if TKA/TKR will be needed in the future.

Arthroscopic Partial Meniscectomy or Total Knee Replacement After Rehabilitation
There were three reports of conversion to APM or TKR after rehabilitation [13,24,35]. The reasons given included increased pain and decreased knee function and satisfaction [13,24]. In addition, patients with high WOMAC physical-function scale scores at six months post-intervention underwent APM or TKR treatment [35]. Therefore, any exacerbation of pain or decline in physical function after rehabilitation may have led to the transition to APM or TKR.

Clinically Important Differences
Cut-offs, CIDs, and MICs are important indices in knee OA, as they are associated with meniscus injury; these were examined in three studies [24,28,35]. In addition, a study on knee OA reported minimal CIDs (MCID) in NRS, SF-36, VAS, and WOMAC scores [42]. Specifically, they reported an NRS of 1.0, an SF-36 (physical function) score of 3.3 points, a VAS of -8.4 to -19.9 mm, and a WOMAC (pain/physical function) score of -9.7/-9.3 points [42]. Moreover, MCII was reported to be -9.9 mm for VAS and -9.1 (-26.0%) points for WOMAC physical function [43]. In contrast, some studies involving middle-aged and older patients with meniscus injuries reporting IKDC scores defined MIC as 10.9 points [44]. On the other hand, the MIC of the PSFS is reported at 2.5 points [28]. Hence, we identified the studies reported in this review that met these definitions; nine met these criteria [13,26,27,29,[32][33][34][35][36]. The MCII is reportedly unaffected by age, duration of disease, or sex, and we believe that this is a useful definition for utilization in clinical practice [45]. Therefore, we believe that the outcomes of knee OA associated with meniscus injuries should be evaluated by KOOS subscales (including pain/symptoms, physical function, ADL, sport/recreation, and QOL) or WOMAC scores, and cut-off values, CIDs, MICs, MCIDs, and MCIIs should be applied in clinical practice.

Clinical Implications
Although this scoping review cannot distinguish and evaluate the degree of meniscus injuries, we believe that ET for knee OA associated with meniscus injuries should consist of a stepwise program of interventions from the acute to the sub-acute phases and finally to the advanced activity phase [35]. In addition, the cutoff values, CID, MIC, MCID, and MCII may have important applications in clinical settings.

Limitations
Our scoping review has some limitations regarding its methodology and interpretation of results. First, the inclusion criteria for this review incorporated medications in addition to rehabilitation; hence, it was impossible to examine the effects of rehabilitation alone. Studies examining MMPRT require cautious interpretation because they cannot present the effects of rehabilitation alone because pharmacotherapy is used for a certain period. Second, because most studies compared APM with rehabilitation, it is unclear whether there was a significant difference before and after the rehabilitation-only intervention. Third, the methods and effectiveness of rehabilitation at the patellofemoral joint (PFJ) are unknown; knee OA affects the PFJ and tibiofemoral joint OA. This review provides an important perspective and points for consideration given that meniscus tears reportedly increase the risk of PFJ OA [49]. Additionally, it is important to note that for tibiofemoral joint OA and PFJ OA, no specific interventional procedure is indicated.

Future research
Further intervention studies are needed regardless of the extent of the meniscus injury as the definition of the MCID is important to generalize the efficacy of rehabilitation alone in treating knee OA associated with meniscus injuries.

Conclusions
Studies on rehabilitation of middle-aged and older patients with knee arthritis associated with meniscus injury had an evidence gap, with differences in intervention duration, methods, and outcomes. Additionally, at short-term follow-up, intervention effects varied across studies. Therefore, the rehabilitation approach should consist of a stepwise program that improves physical function and reduces pain and other symptoms. Furthermore, the follow-up period should be at least one year, which can be considered medium-to-long term, and outcomes should utilize the KOOS and WOMAC scores to improve pain, symptoms, and physical function, as well as ADL and QOL based on the cut-off scores, CID, MIC, MCID, or MCII definitions. Knee OA associated with meniscus injury may be a point of consideration in clinical practice and research as it may lead to a certain number of patients transitioning to surgical repair or replacement in the long term.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.