Mechanical Symptoms and Meniscal Tear: A Reappraisal

Abstract

Objective:

To identify and summarize literature related to the association between mechanical symptoms (catching and locking of the knee), the presence of meniscal tear, and outcomes after arthroscopic surgery.

Design:

We searched PubMed and hand-searched reference lists for relevant articles and selected 38 for analysis.

Results:

Mechanical symptoms appear to have modest sensitivity (ranging 0.32–0.69), specificity (ranging 0.45–0.74) and positive predictive value (ranging 0.75–0.81) for meniscal tear. There is also very little evidence to suggest that those with mechanical symptoms experience better outcomes after arthroscopic surgery.

Conclusion:

Our examination of the literature does not support the hypothesis that mechanical symptoms are related to the presence of meniscal tear or portend better outcomes after arthroscopic surgery.

In 2020, over 400,000 arthroscopic partial meniscectomies (APM) were performed on persons with knee pain, meniscal tear and some degree of osteoarthritic damage in the United States.¹ Clinicians have been taught for decades that in patients with knee pain the diagnosis of meniscal tear is more likely –and the outcomes of APM are better – in the presence of “mechanical symptoms,” such as knee catching, clicking, and locking. While considerable research has been done to address APM efficacy, there has been less careful examination of the definition of “mechanical symptoms,” whether these symptoms are associated with meniscal tear, and whether patients with mechanical symptoms are more likely to improve after APM.^2–6 Here we review literature on the association of mechanical symptoms with meniscal tear and on the relationship between mechanical symptoms and post-operative outcome.

To address these questions, we searched PubMed for human studies using the term “mechanical symptoms AND knee AND menisc*”. The initial search yielded 586 papers, which were then manually screened by title, abstract, and full text. We also hand-searched reference lists for relevant articles. We selected 38 publications for this narrative analysis, focused on clinical presentation or treatment of meniscal tear.

Association between meniscal tear and mechanical symptoms

We first examine the relationship between meniscal tear and mechanical symptoms. Mechanical symptoms are commonly defined as the presence of clicking, catching, or locking in the knee.^7–10 This is the definition most often used in the literature, although some authors have included the knee “giving way” as well (Table 1). We measure the association between mechanical symptoms and meniscal tear using sensitivity, specificity, and positive predictive value (PPV). Here the gold standard is presence of meniscal tear during arthroscopy or on MRI. We examine the sensitivity, specificity and PPV of having mechanical symptoms in relation to this gold standard. Sensitivity is the proportion of subjects with meniscal tear who have mechanical symptoms. Specificity is the proportion of subjects with no meniscal tear who do not have mechanical symptoms. PPV represents the proportion of subjects with mechanical symptoms who have meniscal tear. PPV is prevalence dependent.¹¹ We illustrate sensitivity, specificity and the prevalence dependence of PPV in Table 2.”

Table 1.

Sensitivity, specificity, and positive predictive value (PPV) of mechanical symptoms for meniscal tear.

Author	Year	Definition of mechanical symptoms	Measure of mechanical symptoms	Diagnostic standard for meniscal tear	Sample size	Prevalence of meniscal tear	Sensitivity	Specificity	PPV
Noble	1975	Locking	Self-reported, scale not reported by authors	Arthroscopy	113	1.00	0.32	N/A	N/A
Aglietti	1985	Locking	Physician-reported upon physical examination	Arthroscopy	100	1.00	0.59	N/A	N/A
Friedman	1987	Locking, clicking, catching, buckling	Self-reported, scale not reported by authors	Arthroscopy	42	1.00	0.50	N/A	N/A
Niu	2011	Clicking, giving way, catching, locking	Self-reported, using the Meniscal Symptom Index	MRI	300	0.40	0.47–0.69	0.5–0.75	N/A
Sihvonen	2016	Locking, catching	Self-reported using an adapted version of the locking domain of Lysholm knee score	Arthroscopy	932	1.00	0.63	N/A	N/A
Pihl*	2017	Catching, locking	Self-reported frequency of symptoms, ranging from “never” to “daily”	Arthroscopy	620	1.00	0.51	N/A	N/A
Hare	2017	Catching	Self-reported on 5-point Likert scale of severity	MRI	199	1.00	0.41	N/A	N/A
MacFarlane	2017	Clicking, catching, popping, giving way, locking, swelling, pain with pivoting or twisting	Self-reported frequency of symptoms in the last week, ranging from “never” to “several times per day”	MRI	227	1.00	0.49	N/A	N/A
Pihl*	2019	Catching, locking, extension deficit	Self-reported, using KOOS symptom subscale	Arthroscopy	566	1.00	0.68	N/A	N/A
Thorlund*	2019	Catching, locking, extension deficit	Self-reported, using KOOS symptom subscale	Arthroscopy	817	0.78	0.53	0.36	0.75
Sihvonen	2020	Catching, locking	Self-reported using the locking domain of the Lysholm knee score	Arthroscopy	146	1.00	0.47	N/A	N/A
Farina	2021	Catching, clicking, pain with pivoting or twisting	Self-reported, using KOOS symptom subscale items	Arthroscopy	565	0.79	0.65	0.45	0.81
MacFarlane	2021	Clicking, catching, popping, intermittent locking, giving way, swelling	Self-reported frequency of symptoms in the last week, ranging from “never” to “several times per day”	MRI	287	1.00	0.48–0.67	N/A	N/A

^*Patients in these studies are from the same cohort.

Table 2.

An example of two populations with varying prevalence of meniscal tear, assuming that the sensitivity and specificity of mechanical symptoms for meniscal tear are both 0.80.

A. Meniscal tear prevalence of 40%					B. Meniscal tear prevalence of 20%
	+ tear	− tear	total			+ tear	− tear	total
+mech sx	32	12	44		+mech sx	16	16	32
− mech sx	8	48	56		−mech sx	4	64	68
total	40	60	100		total	20	80	100
PPV = 73% Sensitivity and specificity = 0.80					PPV = 50% Sensitivity and specificity = 0.80

Studies using operative findings as the gold standard suggest the sensitivity of mechanical symptoms for meniscal tear in cohorts with surgically confirmed meniscal tear is moderate at best. In an observational cohort study of persons undergoing arthroscopic surgery, Farina and colleagues found that the sensitivity of mechanical symptoms for meniscal tear was 0.65.⁷ In the FIDELITY trial of 146 subjects with surgically confirmed meniscal tear, Sihvonen et al found that only 47% reported mechanical symptoms and APM had no added benefit over sham surgery to alleviate these symptoms.⁹ The finding that APM was unable to relieve these symptoms more than sham surgery illustrates clearly the discordance between meniscal tear and mechanical symptoms. Dervin at el reported in 2001 that among 152 patients with knee pain who were evaluated for unstable meniscal tear and underwent arthroscopy, mechanical symptoms were not reliable predictors of meniscal tear.¹² In seven more studies comprising over 1000 patients undergoing arthroscopy, mechanical symptoms had sensitivity ranging from 0.32–0.68 for arthroscopically-confirmed meniscal tear (Table 1).^{8, 13–18}

In studies that utilized MRI to confirm the presence of meniscal pathology, the sensitivity of mechanical symptoms for meniscal tear was similarly modest. It has been shown previously that many individuals with meniscal pathology on MRI experience no pain or mechanical symptoms.¹⁹ In a cohort of 199 middle-aged patients with MRI-verified degenerative meniscal tear, mechanical symptoms had sensitivity of 0.41.²⁰ Similarly, MacFarlane et al noted that mechanical symptoms had sensitivity of 0.49 for meniscal tear in a trial cohort of 227 middle-aged and elderly patients with tears confirmed on MRI.²¹ Niu and colleagues reported that various individual mechanical symptoms (e.g. catching, clicking, locking, giving way) had sensitivities of 0.47–0.69 in a sample of subjects with MRI-verified meniscal tear (Table 1).²²

The specificity of mechanical symptoms has been examined less frequently than sensitivity, but the findings suggest modest specificity. In three studies published between 2011 and 2021, the specificity of mechanical symptoms ranged from 0.36–0.74.^{7, 13, 22} Positive predictive value is prevalence-dependent and therefore less valuable for assessing performance across studies (Table 2). With that caveat, two studies with meniscal tear prevalence of 0.78 and 0.79 reported positive predictive values of mechanical symptoms of 0.75 and 0.81, respectively (Table 1).^{7, 13}

Despite the consistent finding that mechanical symptoms have limited sensitivity and specificity for meniscal tear, expert opinion articles and reviews have equated mechanical symptoms with meniscal tear and/or supported the use of APM in patients with mechanical symptoms for decades.^23–36 For example, in 1973, JA Nicholas wrote that the “diagnosis of meniscal tear can be made both by a history of “locking” or “slipping” as well as by physical examination.”³⁰ This trend continued through 2002, when Greis et al wrote that “degenerative tears of the menisci tend to occur in older patients (>40 years), frequently with… mechanical symptoms.”³³ More recently, Chirichella and colleagues stated in 2019 that “in the chronic [meniscal tear] setting, patients might complain of knee pain associated with intermittent swelling and mechanical symptoms.”²⁹ A review of diagnostic and management strategies for meniscal tear published in 2020 reported that “patients already diagnosed with severe osteoarthritis on plain X-rays and exhibiting meniscal mechanical symptoms can be safely assumed to have a meniscal tear.”³⁶ Despite increasing evidence showing the limited sensitivity and specificity of mechanical symptoms for meniscal tear, expert guidance to the contrary has persisted.

Association between mechanical symptoms and arthroscopic partial meniscectomy outcomes

Another principle taught to generations of clinicians in the last half-century is that patients who have mechanical symptoms are most likely to benefit from arthroscopic partial meniscectomy. Several published expert opinions advise using mechanical symptoms as an indication of favorable candidacy for arthroscopy. For example, a 1995 review of 103 geriatric patients undergoing arthroscopy found that 43% had mechanical symptoms.³⁷ 96% of those with mechanical symptoms were found to have a good or excellent result after APM, compared to only 42% of those without mechanical symptoms reporting the same. Similarly, a 1997 review of arthroscopic treatment of knees with degenerative damage stated that “mechanical symptoms, suggestive of unstable meniscal tears or loose bodies, correlate with good postoperative results.”²⁸

This advice has persisted. A 2012 survey of orthopedic surgeons ascertained clinical variables the respondents used to decide whether to recommend arthroscopic partial meniscectomy.³⁸ Mechanical symptoms were not included as a clinical variable in the survey because “it was believed that there would be nearly uniform agreement that patients with such symptoms would require surgery.”

However, in the first decade of the twenty-first century, studies emerged that challenge the traditional understanding that mechanical symptoms are associated with meniscal tear and portend better outcomes after arthroscopy.^{12, 39} A recent analysis of MeTeOR trial data indicated that those with mechanical symptoms did not experience greater pain relief from APM or physical therapy than those without.¹⁰ In 2018, Pihl et al reported that younger patients (mean age 31) with mechanical symptoms had had larger improvements in KOOS4 score than their counterparts without mechanical symptoms after APM (adjusted mean difference 10.5, 95% CI 4.3 to 16.6).⁴⁰ However, this difference in symptom relief between those with and without mechanical symptoms was not observed in older patients (mean age 54).⁴⁰ The authors posited that older patients often had degenerative tears and cartilage defects reflecting early stage osteoarthritis that could have caused their mechanical symptoms, while younger patients with otherwise normal knees likely had isolated meniscal tears causing their mechanical symptoms.

Gauffin et al found in an as-treated analysis of a randomized controlled trial of APM that patients with mechanical symptoms who were treated with surgery experienced less pain relief from the procedure than those without mechanical symptoms.³ Initial findings from the same trial published in 2014 indicated that neither age nor mechanical symptoms affected the outcome of APM.⁴¹

In a 2016 cohort study of APM outcomes, Sihvonen et al reported that patients with mechanical symptoms self-reported less satisfaction following surgery (61% vs 75%, multivariable adjusted risk ratio [RR] 0.84; 95% confidence interval [CI] 0.76, 0.92), and less self-rated improvement (RR 0.91; 95% CI 0.85, 0.97) than those without mechanical symptoms.¹⁵ No statistically significant difference was found in the change in WOMET score or numerical rating scale pain between those with and without mechanical symptoms after surgery. In a study published in 2008, Kirkley et al found no benefit conferred by knee arthroscopy in a subgroup analysis among patients with mechanical symptoms.³⁹

Despite these reports, practice guidelines have continued to state that patients with mechanical symptoms experience better outcomes after APM, and posited that mechanical symptoms are a good indicator of meniscal tear.^{26, 27, 32–35, 42, 43} The National Institute of Health and Care Excellence recommends against arthroscopic lavage and debridement in patients with knee osteoarthritis, unless they have a “clear history of mechanical locking.”⁴⁴ The European Society of Sports Traumatology, Knee Surgery, and Arthroscopy (ESSKA) recommends non-operative treatment for meniscal tear, except in the case of “considerable mechanical symptoms.”⁴⁵ On the other hand, The American Academy of Orthopedic Surgeons does not comment specifically on the use of mechanical symptoms as an indication for arthroscopic partial meniscectomy.⁴⁶

It appears, then, that practice has lagged behind science. The best evidence available suggests that mechanical symptoms are not useful clues to the diagnosis of symptomatic meniscal tear, but rather are more closely related to osteoarthritis than meniscal tear. Evidence also suggests that these symptoms fail to identify patients more likely to benefit from APM.

Patients with mechanical symptoms or meniscal tear can experience significant functional limitations that may impact their ability to exercise, which is often the first line of treatment for these patients. Current studies rarely address the severity of mechanical symptoms, and the studies included here report mechanical symptoms as a dichotomized variable (i.e. the presence or absence of symptoms). Experienced clinicians generally tailor the choice of treatment for these patients to the level of functional limitation, patient-reported quality of life, and clinical findings. We recommend that this more contemporary view of mechanical symptoms be integrated into practice guidelines, teaching materials, texts, and conversations between physicians and their patients.