Meniscus—Friend or foe: Epidemiologic observations and surgical implications


  • Jeffrey N. Katz,

    Corresponding author
    1. Brigham and Women's Hospital, Harvard Medical School, and Harvard School of Public Health, Boston, Massachusetts
    • Orthopaedic and Arthritis Center for Outcomes Research, Brigham and Women's Hospital, OBC-4, 75 Francis Street, Boston, MA 02115
    Search for more papers by this author
  • Scott D. Martin

    1. Brigham and Women's Hospital, and Harvard Medical School, Boston, Massachusetts
    Search for more papers by this author

The meniscus has long been appreciated as a source of knee symptoms and disability. In fact, through much of the 20th century, total meniscectomy was performed relatively frequently for patients with persistent, disabling mechanical knee symptoms. The procedure relieved mechanical symptoms but at a cost: total meniscectomy was associated with accelerated degenerative radiographic changes, establishing the importance of the meniscus in the pathogenesis of osteoarthritis (OA) (1–3). Biomechanical studies clarified these clinical observations. The meniscus distributes substantial mechanical load throughout the tibiofemoral joint surfaces that would otherwise be transmitted directly and focally to the weight-bearing portions of the femur and tibia. Damage to or resection of the meniscus disrupts this load-sharing function, leading to excess loading of cartilage and subchondral bone and the attendant changes of OA (1, 4–6).

These concerns assumed public health importance with the advent of 2 key technologies, arthroscopy and magnetic resonance imaging (MRI). The work of Watanabe and other investigators in the 1960s led to the introduction of arthroscopy in the 1970s and its widespread dissemination by the 1980s (7). The arthroscope made meniscal surgery much simpler and less costly. Widespread application of arthroscopic meniscectomy required noninvasive diagnostic techniques to facilitate recognition of meniscal tears. MRI filled this need, with sensitivities and specificities for meniscal tear on the order of 90% as compared with a surgical gold standard (8–10). The combination of MRI and arthroscopy permitted noninvasive diagnosis and minimally invasive treatment of symptomatic meniscal tears (11). By the end of the 20th century, arthroscopic knee surgery had become the most common procedure performed by orthopedic surgeons, with up to 1 million knee arthroscopies performed annually in the US at present (12, 13). The majority of these procedures appear to be performed for a meniscal tear. The lack of a comprehensive ambulatory surgery database in the US precludes more precise estimates of arthroscopy utilization.

Technology development and dissemination often subtly change the spectrum of diagnosed and treated disease. The typical recipient of meniscal surgery prior to the age of arthroscopy and MRI was a younger person with a relatively normal knee joint and an injury-related meniscal tear. Over time, the spectrum shifted to include older individuals with underlying OA and superimposed degenerative meniscal tears. This population is difficult to evaluate, because the vast majority of patients with OA also have meniscal tears, as do one-third of asymptomatic adults (14–16).

The capacity to identify a torn meniscus noninvasively, and to address it minimally invasively, has emboldened surgeons and patients to opt for arthroscopic partial meniscectomy. The efficacy of arthroscopic partial meniscectomy in the setting of OA has received surprisingly little study. Herrlin and colleagues performed the only randomized controlled trial addressing this question and found no difference in function or pain between patients treated with arthroscopic partial meniscectomy or with a nonoperative regimen for symptomatic meniscal tear and concomitant OA (17).

Thus, we know relatively little about the efficacy of arthroscopic partial meniscectomy in patients with concomitant OA and even less about the effects of meniscal damage and meniscal resection on the underlying degenerative process. As noted above, total meniscectomy is clearly associated with degenerative changes and accelerated OA (3). Observations over the past 2 decades have shown that partial meniscectomy is also associated with accelerated OA (18). These findings have led to the development of surgical approaches that preserve or replace meniscal tissue, such as meniscal repair and meniscal allograft implantation (5, 19, 20). Partial meniscectomy has evolved in the same manner, with increasing emphasis on trimming only the unstable portion of the tear and leaving the stable portion intact (4).

It is unclear whether the osteoarthritic changes that follow total or partial meniscectomy arise from the surgery itself or from the underlying condition. Specifically, does the resection of meniscal tissue lead to joint instability and attendant joint damage (as is presumed to occur following total meniscectomy), or is the problem not surgery but the underlying meniscal lesion itself and its associated abnormalities in cartilage, bone, and synovium? This question cannot be addressed simply by evaluating the outcome of patients who have undergone meniscal surgery, because the role of the underlying meniscal lesion and the surgical procedure cannot be disentangled in such studies.

In this issue of Arthritis & Rheumatism, Englund et al address this question with an elegant nested case– control study (21). The parent cohort included individuals ages 50–79 years participating in the Multicenter Osteoarthritis (MOST) Study, which was designed to identify predictors of OA incidence and progression in a large community-based sample. The nested case–control study was restricted to 1,863 MOST Study subjects who were at risk but initially free of radiographic OA at the baseline evaluation. After 30 months of followup, definite OA (defined as Kellgren/Lawrence [22] grade 2) developed in 121 of these individuals, who had readable MRI examinations at baseline. These patients served as cases and were compared with 294 control subjects who were also free of OA at the outset but in whom OA did not develop. The authors examined the baseline MRIs for meniscal changes and observed that 3-fold more cases than controls had meniscal damage at baseline (54% versus 18%).

Thus, among older patients who did not have knee surgery of any sort, the presence of meniscal damage is associated with the subsequent development of OA. It could be argued that much of the population under study already had OA at the time of study entry, but that it had not yet become evident on radiographs. The authors addressed this issue by examining a subset of patients with no cartilage findings at baseline and confirmed in this group as well that the presence of baseline meniscal damage predicted the subsequent development of cartilage lesions.

This study establishes that meniscal damage is a risk factor for the development of radiographic OA. Returning to the conundrum of whether the OA that follows arthroscopic partial meniscectomy is attributed to partial meniscectomy or to the underlying meniscal lesion, the work of Englund et al indicates that the meniscal lesion itself likely plays a part. How large a part? The relative contributions from the meniscal lesion, from associated joint abnormalities, and from arthroscopic partial meniscectomy surgery can only be resolved with a randomized controlled trial of arthroscopic partial meniscectomy versus nonoperative therapy among patients with symptomatic meniscal tear and concomitant OA. Such a randomized controlled trial would permit us to estimate the counterfactual: what would have happened if the same patient who we observe several years after arthroscopic partial meniscectomy had not undergone surgery?

The Englund study prompts several suggestions for research priorities and clinical management. First, these data support continued work on strategies to preserve meniscal tissue in the management of meniscal tear. Rigorous studies are needed to address both the symptomatic response to meniscus-sparing surgery and the effects of these surgeries on the development and progression of OA. The results reported by Englund and colleagues warn that control observations will be critical to testing the hypothesis that meniscus-preserving surgery slows the development of OA.

A second implication is that ongoing research is needed on the determinants of meniscal degeneration, because a damaged meniscus appears to be a potent risk factor for OA. In particular, the prevention of knee injury may have an even more important role in reducing the population burden of OA than was previously envisioned. Injury prevention is typically oriented toward reducing the growing number of anterior cruciate ligament tears, because these lesions clearly lead to accelerated OA if they are not managed surgically. If meniscal tears are also potent risk factors for accelerated OA, knee injury prevention assumes an even greater public health priority.

Patients will continue to present with symptoms suggestive of meniscal injury. Those with compelling mechanical symptoms and functional disability will continue to undergo arthroscopic partial meniscectomy. The work of Englund and colleagues adds to the call for critical evaluation of the indications for and rates of surgery for meniscal lesions in middle-aged patients, to ensure we are doing more good than harm.