Meniscal Surgery: Risk of Radiographic Joint Space Narrowing Progression and Subsequent Knee Replacement Data from the Osteoarthritis Initiative 1

Transcription

1 This copy is for personal use only. To order printed copies, contact Bashir Zikria, MD, MSc Nima Hafezi-Nejad, MD, MPH Frank W. Roemer, MD Ali Guermazi, MD, PhD Shadpour Demehri, MD Meniscal Surgery: Risk of Radiographic Joint Space Narrowing Progression and Subsequent Knee Replacement Data from the Osteoarthritis Initiative 1 Purpose: To investigate the risk of radiographic joint space narrowing (JSN) progression evaluated in subjects with and those without underlying osteoarthritis at baseline and knee replacement (KR) associated with meniscal surgery in subjects with and those without a reported history of preceding knee trauma. Original Research n Musculoskeletal Imaging 1 From the Department of Orthopedic Surgery (B.Z.) and Russell H. Morgan Department of Radiology and Radiological Sciences (N.H.N., S.D.), Johns Hopkins University School of Medicine, 601 N Caroline St, JHOC 5165, Baltimore, MD 21287; and Quantitative Imaging Center, Department of Radiology, Boston University School of Medicine, Boston, Mass (F.W.R., A.G.). Received January 14, 2016; revision requested March 22; revision received June 8; accepted July 11; final version accepted August 18. Address correspondence to S.D. ( demehri2001@yahoo.com). A.G. and S.D. contributed equally to this work. q RSNA, 2016 Materials and Methods: Results: Conclusion: The HIPAA-compliant protocol was approved by the institutional review boards of the participating centers. Written informed consent was obtained from all participants. Subjects who underwent meniscal surgery with a preceding knee trauma at baseline (n = 564) and those without (n = 147) were drawn from the Osteoarthritis Initiative cohort (n = 4796). Radiographic JSN progression was evaluated by using Osteoarthritis Research Society International grading (progression in 1st-, 2nd-, 3rd-, 4th-, 6th-, or 8th-year radiographic findings compared with baseline). KR was assessed up to the 9th year of study (days passed from inclusion to KR or last follow-up). Cox hazard analysis was used to extract the adjusted hazard ratios (HRs) with adjustments for baseline age, sex, body mass index, physical activity, symptoms, and radiographic osteoarthritis features (Kellgren and Lawrence [KL] grade). Meniscal surgery with a history of preceding knee trauma was not associated with radiographic progression of JSN (adjusted HR, 0.91 [95% confidence interval {CI}: 0.78, 1.07]) or KR (adjusted HR, 1.02 [95% CI: 0.79, 1.34]; P =.854). However, meniscal surgery without a history of preceding knee trauma was associated with radiographic progression of JSN (adjusted HR, 1.27 [95% CI: 1.00, 1.63]) and KR (adjusted HR, 2.09 [95% CI: 1.52, 2.89]; P,.001). Additionally, there was a higher risk of KR in subjects with radiographic KL grade of less than 2 (adjusted HR, 6.97 [95% CI: 3.56, 13.64]; P,.001) at baseline in comparison with KL grade of at least 2 (adjusted HR, 1.76 [95% CI: 1.22, 2.54]; P,.05). In contrast to subjects without a reported preceding trauma, meniscal surgery is not independently associated with increased risk of radiographic JSN progression and KR in patients with a reported preceding trauma. q RSNA, 2016 Online supplemental material is available for this article. Radiology: Volume 282: Number 3 March 2017 n radiology.rsna.org 807

2 Meniscal tears are among the most common types of structural knee injury (1), which can either occur after an episode of knee trauma during sports or in older individuals as a degenerative process without a traumatic onset (2). Overall, there is an increasing trend in the incidence of meniscal tears, which can potentially jeopardize the future function and force absorption of the affected knee (2,3). There are two main categories for the management of meniscal tears: conservative management and surgical treatment, most of which are partial meniscectomies. In a few cases, complete meniscectomy, meniscus repair, and meniscal allograft transplantation are also performed (3). Meniscal surgery, which is aimed at removing or repairing torn meniscal fragments, is the most common orthopedic surgical treatment performed in the United States (4). The choice of treatment typically depends on the patient characteristics and the type of the tear (3,5). Several studies demonstrated that the chronic function and the risk of Advances in Knowledge nn In the presence of reported traumatic onset of symptoms, meniscal surgery was not associated with 8-year radiographic joint space narrowing (JSN) progression (adjusted hazard ratio [HR], 0.92 [95% confidence interval {CI}: 0.77, 1.09]; P =.321) or 9-year incidence of knee replacement (KR) (adjusted HR, 1.02 [95% CI: 0.79, 1.34]; P =.854) when adjusted for known risk factors of osteoarthritis (OA) progression among the 4796 participants in the cohort. nn In the absence of reported traumatic onset of symptoms, meniscal surgery is associated with radiographic JSN progression (adjusted HR, 1.45 [95% CI: 1.12, 1.89]; P,.001) and KR (adjusted HR, 2.09 [95% CI: 1.52, 2.89]; P,.001) when adjusted for known risk factors of OA progression. secondary knee osteoarthritis (OA) are closely associated with the amount of residual meniscal tissue after meniscal surgery (6). It is clear that total meniscectomy, especially in long-term evaluations, is associated with higher risk of knee replacement (KR) (7,8). However, literature regarding such associations of all meniscal surgeries, most of which currently include partial meniscectomy and a higher risk for KR, is scarce, to our knowledge (9). In addition, in the setting of traumatic meniscal tears and posttraumatic baseline knee structural damage, it is unclear how meniscal surgery changes the natural course of the injured knee to develop advanced OA. By using the Osteoarthritis Initiative (OAI) database, including the longitudinal evolution of joint space narrowing (JSN) for up to the 8th year and KR for up to the 9th year, we aimed to describe the natural history of subjects with a reported history of meniscal surgeries. Thus, the purpose of our study was to investigate the risk of radiographic JSN progression (evaluated in subjects with underlying OA at baseline and those without) and KR associated with meniscal surgery in subjects with a reported history of preceding knee trauma and those without. Materials and Methods The data for our study were obtained and analyzed from the OAI cohort. OAI is a longitudinal, prospective, observational cohort study of knee OA that is conducted by personnel at four clinical centers and at one coordinating center in Columbus, Ohio; Baltimore, Maryland; Pittsburgh, Pennsylvania; and Pawtucket, Rhode Island. The primary goal of the OAI cohort was to establish Implication for Patient Care nn In subjects with meniscal tears without a reported history of preceding knee trauma, meniscal surgery may be associated with an increased risk of radiographic JSN progression, especially in those with a radiographic Kellgren and Lawrence grade of 0 or 1. a natural history database of knee OA, including clinical evaluation data, radiologic images, and a biospecimen repository. From February 2004 to May 2006, 4796 individuals who had OA or were at risk for knee OA were recruited and were followed up for incidence and progression of knee OA. Inclusion criteria comprised knee symptoms in the preceding year, obesity, knee injury, knee surgery, family history of OA, Heberden nodes, and repetitive knee-bending activities, as well as frequent knee symptoms and radiographic tibiofemoral OA features. Exclusion criteria comprised rheumatoid arthritis or inflammatory arthritis, bone-on-bone baseline radiographic findings, baseline bilateral KRs, inability to undergo 3.0-T magnetic resonance (MR) imaging, positive pregnancy test results, inability to provide blood samples, and other interfering comorbid conditions. The Health Insurance Portability and Accountability Act compliant protocol was approved by the institutional review boards of the participating centers. Annual radiographic Published online before print /radiol Content code: Radiology 2017; 282: Abbreviations: BMI = body mass index CI = confidence interval HR = hazard ratio JSN = joint space narrowing KL = Kellgren and Lawrence KR = knee replacement OA = osteoarthritis OAI = Osteoarthritis Initiative OARSI = Osteoarthritis Research Society International PASE = Physical Activity Scale for the Elderly WOMAC = Western Ontario and McMaster Universities Arthritis Index Author contributions: Guarantors of integrity of entire study, B.Z., N.H.N., S.D.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, B.Z., N.H.N., A.G., S.D.; clinical studies, all authors; experimental studies, S.D.; statistical analysis, B.Z., N.H.N., S.D.; and manuscript editing, all authors Conflicts of interest are listed at the end of this article. 808 radiology.rsna.org n Radiology: Volume 282: Number 3 March 2017

3 examinations and assessments of KR (and recording the exact date of KR) were performed, and informed consent was obtained from all participants. OAI data sets are available by using the Web portal Our analysis was conducted on the original cohort. From the baseline evaluation, subjects age, sex, and body mass index (BMI) data were included in the analysis. The Physical Activity Scale for the Elderly (PASE) score was used to assess the level of physical activity in each subject (10). The PASE is a reliable, valid, and widely used instrument for the quantification of physical activity in epidemiologic studies, especially among the elderly. Baseline PASE score was used for analysis. The Western Ontario and McMaster Universities Arthritis Index (WOMAC) questionnaire provides a validated and reliable evaluation of patients symptoms by asking questions regarding pain, stiffness, and physical function of the affected joint(s) (11). The WOMAC questionnaire results were assessed in every participant, and the baseline WOMAC total score was used for analysis. The Kellgren and Lawrence (KL) grade was used to demarcate the baseline radiographic signs of OA. For individuals with bilateral knee symptoms, the worst WOMAC scores and KL grades were used for analysis. PASE scores, WOMAC scores, and KL grades from the baseline evaluation were included in the adjusted models. Our previous study on the OAI cohort showed the association between use of analgesics (based on medical inventory forms) and OA progression and KR (12). Thus, we further evaluated the status of analgesic consumption among the subjects studied and confirmed our adjusted results by accounting for the baseline use of analgesics. Study Population Subjects were categorized into three groups according to their history of meniscal surgeries: group 1, subjects who underwent meniscal surgery without a history of preceding knee trauma; group 2, subjects who underwent meniscal surgery with a history of preceding knee trauma; and group 3, the remainder of the cohort. Information used to select these groups was extracted by using two questions in the baseline OAI visits: (a) Did you ever undergo a meniscectomy (where they repaired or cut away a torn meniscus) in your right and/or left knee? and (b) Was this meniscal surgery performed to repair an injury episode? All other subjects who did not report meniscal surgeries were categorized as the remainder of the cohort. Baseline age, sex, BMI, PASE score, WOMAC score, and KL grade were selected by following previous risk prediction models as the major risk factors of OA progression and KR (13 16). Outcome Assessment Two outcomes were selected for this analysis. Radiographic progression of JSN (up to the 96th month or 8th year of the study). Osteoarthritis Research Society International (OARSI) scoring of knee radiographs for JSN from the central OAI readings was used for this purpose (17). In a previous OAI study, the reliability of the central OAI readings of knee radiographs was evaluated and confirmed (KL grades and OARSI scores for JSN) (18). OARSI scoring of knee radiographs for JSN is based on a scale of 0 3, where 0 indicates normal findings and 3 indicates end-stage knee OA. Baseline radiographs were compared with follow-up radiographs (1st-, 2nd-, 3rd-, 4th-, 6th-, and 8thyear radiographs) to detect OA progression by means of JSN progression (Table E1 [online]). Subjects with baseline JSN grade less than 3 and definite JSN increase were classified as having JSN progression. By using the central OAI readings of knee radiographs, definite within-grade or full-grade JSN progression (according to the OARSI atlas) was considered to represent the radiographic progression of JSN. The within-grade increase corresponds to the joint space width that has been definitely narrowed but by less than a whole OARSI score. Details of the OAI centralized readings and the OARSI scoring of knee radiographs by using JSN as an outcome in the OAI cohort can be accessed online at epi-ucsf.org/datarelease/sasdocs/ kxr_sq_bu_descrip.pdf and oai.epi-ucsf.org/datarelease/sasdocs/ Outcomes_descrip.pdf, respectively. KR (up to the 9th year of the study). Only adjudicated episodes of KR were included in this analysis (reports of KR, which were evaluated and confirmed within the OAI cohort). In the OAI cohort, KR episodes were adjudicated by using the following information: presurgical, surgical, or postsurgical radiography report; discharge summary; face sheet or physician attestation with International Classification of Diseases codes; or radiographic findings at the University of California, San Francisco. Further information regarding the definitions and descriptions of the outcomes is also available through the OAI Web portal. Statistical Analysis Comparisons of the baseline characteristics across the study groups were performed by using one-way analysis of variance. Analyzing the risk of JSN progression and KR for each group was performed by using Cox regression. Time in the study was defined by using the year of radiographic evaluation (for JSN progression) and using the days from the time of enrollment to the first KR or the last available follow-up visit (for subjects who did not undergo KR). Crude hazard ratios (HRs) and adjusted HRs were calculated, and stepwise adjustments were performed accordingly by considering other risk factors for OA progression and KR (age, sex, BMI, PASE score, WOMAC score, and KL grade) (16). To evaluate the progression of knee OA, the analysis was repeated by only including subjects with established OA at baseline (radiographic KL grade 2). At baseline evaluation, 1951 subjects were in the KL grade category of less than 2, and 2539 were in the KL grade category of at least 2 (3.6% 6.4% had missing values). Next, the risk for radiographic progression of JSN and KR was presented according to different strata of the baseline KL grade (,2 vs 2). All regression models were assessed for Radiology: Volume 282: Number 3 March 2017 n radiology.rsna.org 809

4 Table 1 Baseline Characteristics of the Study Population, Follow-up Events, and Duration of Study Parameter Meniscal Surgery without Preceding Knee Trauma Meniscal Surgery with Preceding Knee Trauma Remainder of the Cohort P Value No. of subjects Age (y) * ,.001 No. of women 73 (49.6)* 188 (33.3) 2543 (62.2),.001 BMI PASE score ,.001 WOMAC total score ,.001 KL grade ,.001 Outcomes JSN (up to the 96th month or 8th year) 71 of 139 (51.1)* 207 of 530 (39.0) 1116 of 3834 (29.1) KR (up to the 9th year) 44 (29.9)* 84 (14.9) 297 (7.3) Follow-up duration Median radiographic follow-up (y) 6 (1 8) 8 (3 8) 8 (6 8) Median follow-up KR (d) 3130 ( ) 3235 ( ) 3236 ( ) Note. Unless otherwise indicated, data are means 6 standard deviations, and numbers in parentheses are percentages. The difference between the denominators and the total number of subjects in each group corresponds to the subjects with missing values for that particular variable. * P,.05 when comparing the categories of knee trauma and meniscal surgery with a preceding knee trauma. P,.05 when comparing the categories of meniscal surgery with a preceding knee trauma and the remainder of the cohort. P,.05 when comparing the categories of knee trauma and the remainder of the cohort. Numbers in parentheses are interquartile ranges. their goodness of fit by comparing the expected values of the cumulative survival function with the observed events (x 2, 20, P..01). Analyses were performed by using SAS software (University Edition; SAS Institute, Cary, NC), JMP Pro software (version 11; SAS Institute), and SPSS software (version 18; SPSS, Chicago, Ill). A P value less than.05 was considered to represent the significance threshold. Results In all, 2804 (58.5%) female and 1992 (41.5%) male participants with a mean age 6 standard deviation of years (range, years) were included in the cohort. There was a statistically significant difference in the mean age and percentage of female subjects (but not BMI) among the three groups studied (mean age, years for the meniscal surgery without preceding knee trauma group, years for the meniscal surgery with preceding knee trauma group, and years for the remainder of the cohort [P,.001]; percentage of female subjects, 49.7%, 33.3%, and 62.3%, respectively [P,.001]). The subjects who underwent meniscal surgery had significantly higher levels of physical activity (PASE score for meniscal surgery without a preceding knee trauma, ; PASE score for meniscal surgery with a preceding knee trauma, ; and PASE score for the remainder of the cohort, [P,.001]). Likewise, subjects who underwent meniscal surgery had significantly higher WOMAC scores (WOMAC score for meniscal surgery without a preceding knee trauma, ; WOMAC score for meniscal surgery with a preceding knee trauma, ; and WOMAC score for the remainder of the cohort, [P,.001]) and KL grades (KL grade for meniscal surgery without a preceding knee trauma, ; KL grade for meniscal surgery with a preceding knee trauma, ; and KL grade for the remainder of the cohort, [P,.001]) at baseline. When comparing subjects who underwent meniscal surgery without a preceding knee trauma versus those with a preceding knee trauma, differences were much less pronounced than between subjects who had undergone meniscal surgery and those who had not undergone meniscal surgery (Table 1). Up to the 96th month of the study, 1394 subjects had radiographic progression of knee OA (51.1% for the meniscal surgery without preceding knee trauma group, 39.1% for the meniscal surgery with preceding knee trauma group, and 29.1% for the remainder of the cohort; P,.001). Within the 9 years of follow-up for KR, 425 subjects underwent their first KR (29.9%, 14.9%, and 7.3% in each group, respectively; P,.001). Table 1 demonstrates the univariate comparisons of the baseline characteristics among the three groups of the study. Subjects who underwent meniscal surgery without a preceding knee trauma (Fig 1) had 1.27 (95% confidence interval [CI]: 1.00, 1.63) and 2.09 (95% CI: 1.52, 2.89) times higher risk for radiographic progression of JSN and KR in the fully adjusted models, respectively. These significantly higher risks were independent of age, 810 radiology.rsna.org n Radiology: Volume 282: Number 3 March 2017

5 Figure 1 Figure 1: Images in a 53-year-old woman who underwent meniscal surgery without preceding knee trauma. (a) Baseline double-echo steady-state coronal MR image of the right knee shows a small residual medial meniscus (arrow). (b) Baseline frontal knee radiograph shows mild OA (KL grade of 1). (c) After 48 months, a follow-up frontal radiograph of the same knee shows severe OA (KL grade of 4); medial minimum joint space width has decreased from 4 mm to 0 mm. The patient underwent KR in the 6th year of follow-up. Table 2 Risk of Radiographic Progression of Knee OA and KR among Subjects Who Underwent Meniscal Surgery without Preceding Knee Trauma (vs the remainder of the cohort) Model and KL Grade sex, and BMI, as well as PASE score, WOMAC score, and KL grade (Table 2). Similar results were obtained when only subjects with established OA at baseline (radiographic KL grade 2) (Table 2) were included. Subjects who underwent meniscal surgery with a preceding knee trauma (Fig 2) had higher risk of radiographic progression of JSN and KR Radiographic Progression of Knee OA HR P Value HR P Value Unadjusted 2.04 (1.60, 2.60), (3.52, 6.63),.001 KL grade (1.15, 1.90) (2.22, 4.23),.001 Adjusted for age, sex, and BMI 1.96 (1.54, 2.50), (3.45, 6.53),.001 KL grade (1.17, 1.93) (2.32, 4.43),.001 Adjusted for age, sex, BMI, PASE score, 1.80 (1.41, 2.30), (2.86, 5.43),.001 and WOMAC score KL grade (1.13, 1.87), (2.14, 4.09),.001 Adjusted for age, sex, BMI, PASE score, 1.27 (1.00, 1.63) (1.52, 2.89),.001 WOMAC score, and KL grade KL grade (0.99, 1.641) (1.46, 2.81),.001 Note. Numbers in parentheses are 95% CIs. in the unadjusted models (HR, 1.44 [95% CI: 1.24, 1.67] and 2.17 [95% CI: 1.70, 2.77], respectively). However, this significant risk diminished after adjustment for KL grade (HR, 0.91 [95% CI: 0.78, 1.07] and 1.02 [95% CI: 0.79, 1.34], respectively). Results of Table 3 suggest that subjects who underwent meniscal surgery with preceding knee trauma had no significantly increased KR risk for radiographic progression of JSN or KR after adjusting for KL grade (HR, 0.91 [95% CI: 0.78, 1.07] [P =.321] and 1.02 [95% CI: 0.79, 1.34], respectively [P =.854]). Similar results were obtained when only subjects with established OA at baseline (radiographic KL grade 2) (Table 3) were included. We further analyzed age at the first meniscal surgery with regard to outcomes. Among subjects who underwent knee trauma, the mean age at the first meniscal surgery was years versus years for subjects with and those without radiographic JSN progression, respectively (P =.219), and years versus years for subjects who underwent KR and those who did not, respectively (P =.074). Among subjects who underwent meniscal surgery with a preceding knee trauma, the mean age at the first meniscal surgery was years versus for subjects with and those without radiographic JSN progression, respectively (P =.685), and years versus years for subjects who underwent KR and those who did not, respectively (P =.581). Radiology: Volume 282: Number 3 March 2017 n radiology.rsna.org 811

6 Figure 2 Figure 2: Images in a 69-year-old man who underwent meniscal surgery with preceding knee trauma. (a) Baseline double-echo steady-state coronal MR image of the right knee shows a small residual medial meniscus (arrow) and superficial thinning of the medial tibial plateau. (b) Baseline frontal knee radiograph shows moderate OA (KL grade of 3) with definite JSN and small marginal diffuse osteophytes. (c) After 48 months, follow-up frontal radiograph of the same knee shows slight worsening in radiographic OA features with a slight increase in osteophyte formation but still a KL grade of 3; medial minimum joint space width has not decreased (1.5 mm at both baseline and 48th-month follow-up). The patient did not undergo KR surgery by the 9th year of follow-up. Table 3 Risk of Radiographic Progression of Knee OA and KR among Subjects Who Underwent Meniscal Surgery with Preceding Knee Trauma (vs the remainder of the cohort) Model and KL Grade We also evaluated the use of analgesics among the study groups. In this regard, significant differences existed between the study groups at baseline evaluation (use of analgesics, 30.8% for knee trauma, 28.7% for meniscal surgery with a preceding knee trauma, and Radiographic Progression of Knee OA HR P Value HR P Value Unadjusted 1.44 (1.24, 1.67), (1.70, 2.77),.001 KL grade (0.90, 1.23) (1.13, 1.87),.001 Adjusted for age, sex, and BMI 1.60 (1.37, 1.87), (2.02, 3.36),.001 KL grade (0.97, 1.34) (1.29, 2.18),.001 Adjusted for age, sex, BMI, PASE score, 1.51 (1.30, 1.77), (1.79, 3.00),.001 and WOMAC score KL grade (0.94, 1.31) (1.22, 2.07).001 Adjusted for age, sex, BMI, PASE score, 0.91 (0.78, 1.07) (0.79, 1.34).854 WOMAC score, and KL grade KL grade (0.76, 1.07) (0.75, 1.29).929 Note. Numbers in parentheses are 95% CIs. 23.3% for the remainder of the cohort; P =.003). Nevertheless, the adjusted HRs remained the same when further adjustment was made for the baseline use of analgesics (adjusted HR for knee trauma, 1.27 [95% CI: 1.00, 1.63) for JSN progression and 2.09 [95% CI: KR 1.52, 2.89] for KR; adjusted HR for meniscal surgery with a preceding knee trauma, 0.91 [95% CI: 0.78, 1.07] for JSN progression and 1.02 [95% CI: 0.79, 1.34] for KR). Next, we investigated the relative risk for radiographic JSN progression and KR for the subjects who underwent knee trauma versus those with a preceding knee trauma. In comparison with subjects with a preceding injury, subjects without a preceding trauma had a higher risk of radiographic JSN progression and KR, even in the fully adjusted model (HR, 1.41 [95% CI: 1.06, 1.86] and 1.89 [95% CI: 1.28, 2.79], respectively; Table 4). Figures E1 and E2 (online) show the Kaplan-Meier curves used to compare the study groups in terms of JSN progression and KR. Finally, we investigated the role of baseline KL grade in the association between knee injury versus surgery with a preceding knee injury (Table 5). In the fully adjusted model, subjects who underwent knee trauma had higher risk for 812 radiology.rsna.org n Radiology: Volume 282: Number 3 March 2017

7 KR if they had a KL grade of less than 2 (6.97 times higher risk [95% CI: 3.56, 13.64]) rather than a KL grade of at least 2 (1.76 times higher risk [95% CI: 1.22, 2.54]; P value for the interaction,.05). In the KL grade stratified results (,2 vs 2), meniscal surgery in subjects with a preceding knee trauma had no additional risk for KR. In other words, the risk of future KR was 6.97 (95% CI: 3.56, 13.64) times higher in a subject with KL grade of less than 2 who underwent knee trauma (adjusted for age, sex, BMI, PASE score, and WOMAC score and compared with a subject with similar KL grade, 2). Meanwhile, the risk of future KR was only 1.76 (95% CI: 1.22, 2.54) times higher in a subject Table 4 with KL grade of at least 2 who underwent knee trauma (adjusted for age, sex, BMI, PASE score, and WOMAC score and compared with a subject with similar KL grade 2). Thus, baseline KL grade modified the effect size (approximately the amount of additional risk) associated with meniscal surgery without a preceding knee trauma with regard to the risk of future KR. Additional details regarding the incidence of JSN and knee replacement are presented in Tables E1 and E2 (online). Discussion Our study demonstrated evidence for a long-term increase in progression of Risk of Radiographic Progression of Knee OA and KR with Regard to Meniscal Surgery without a Preceding Knee Trauma versus with a Preceding Knee Trauma Model Radiographic Progression Features of Knee OA HR P Value HR P Value Unadjusted 1.41 (1.07, 1.85) (1.53, 3.17),.001 Adjusted for age, sex, and BMI 1.31 (0.99, 1.73) (1.24, 2.63).002 Adjusted for age, sex, BMI, PASE score, 1.30 (0.99, 1.72) (1.16, 2.50).006 and WOMAC score Adjusted for age, sex, BMI, PASE score, WOMAC score, and KL grade 1.41 (1.06, 1.86) (1.28, 2.79).001 Note. Numbers in parentheses are 95% CIs. KR radiographic JSN and KR among subjects who underwent meniscal surgery (with or without a preceding knee injury), compared with the remainder of the OAI cohort. After adjustment for the other baseline covariates (including radiographic KL grades), only patients who underwent meniscal surgery without a preceding knee trauma retained a higher risk of radiographic progression of JSN and KR. Because of the significant baseline differences in the study groups, we adjusted the results of our analysis for the baseline parameters, as well. Nevertheless, residual confounding factors can be best addressed in future studies with a randomized clinical trial design. A potential alternative explanation for these results may be that the without a preceding trauma group is simply manifesting a later (or more severe) disease process, making radiographic progression and KR more likely to occur, but without a true causal association with the meniscal surgery. Our conclusion of association, on the basis of the statistical modeling, depends largely on model fit and whether the model adequately captures the relevant variables of risk. While we assessed the goodness of fit of our models, we could not systematically exclude the possibility of missing and unmeasured covariates that could drive the association, rather than the surgery itself. Our results are in line with the previous Table 5 Risk for KR among Subjects Who Underwent Meniscal Surgery without Preceding Knee Trauma or with Preceding Knee Trauma (vs the remainder of the cohort), Stratified according to Baseline Radiographic KL Grade Group and Model KL Grade, 2 P Value for KL Grade, 2 vs Remainder of the Cohort KL Grade 2 Without preceding knee injury Unadjusted 6.97 (3.62, 13.41)*, (1.20, 2.49).003 Adjusted for age, sex, and BMI 7.84 (4.03, 15.27)*, (1.23, 2.56).002 Adjusted for age, sex, BMI, PASE score, and WOMAC score 6.97 (3.56, 13.64)*, (1.22, 2.54).002 With preceding knee injury Unadjusted 0.86 (0.38, 1.97) (0.80, 1.35).763 Adjusted for age, sex, and BMI 1.17 (0.51, 2.72) (0.88, 1.56).269 Adjusted for age, sex, BMI, PASE score, and WOMAC score 1.13 (0.48, 2.62) (0.86, 1.51).375 P Value for KL Grade 2 vs Remainder of the Cohort Note. Data are HRs. Numbers in parentheses are 95% CIs. At baseline evaluation, 1951 subjects were in the KL grade, 2 category, and 2539 were in the KL grade 2 category (3.6% 6.4% had missing values). * P value for the interaction of KL grade (,2 vs 2) is,.05 for meniscal surgery without preceding knee injury. Radiology: Volume 282: Number 3 March 2017 n radiology.rsna.org 813

8 observation that in subjects with degenerative meniscal tears without a preceding trauma, the clinical outcomes after partial meniscectomy were not superior to those after a sham surgical procedure (19). Therefore, our results are in agreement with those of the studies in which the risks associated with surgical management of degenerative meniscal tears were highlighted in subjects without a preceding trauma episode (1,5,20), in which KR was considered to be the ultimate outcome for patients with knee OA (13,15,21). Results of our initial univariate analysis demonstrated subjects symptoms and level of physical activity to be worse for patients who underwent meniscal surgery (with or without a reported history of preceding knee trauma) compared with the remainder of the cohort at baseline. The higher baseline radiographic KL grade in subjects who underwent meniscal surgery may have been due to underlying cartilage damage or concurrent biomechanical abnormalities (22,23). These findings are consistent with those in previous literature, as the presence of meniscal tears that require surgery is commonly associated with knee pain and cartilage damage (24,25). Meniscal surgery in these subjects (with a history of preceding knee trauma) was not associated with increased risk of KR in our study, even in those without a severe radiographic feature of OA (radiographic KL grade, 2) (26). However, these findings do not translate to the safety of meniscal surgery, even with a reported history of preceding knee trauma. Previous MR studies have shown a tendency toward poor clinical outcome after arthroscopic partial meniscectomy (27). Randomized controlled trials have also shown the nonsuperiority of arthroscopic meniscectomy in comparison with supportive therapy or sham surgery (19,28,29). In fact, given the increasing trend in the number of arthroscopic surgeries performed in the United States (compared with Europe) (30) and other potential complications of arthroscopy, including deep venous thrombosis, pulmonary emboli, and infection (in 2.8% 7.6% of overall procedures) (31,32), the higher incidence of subsequent KR in subjects without a reported history of preceding knee trauma in our study should not be overlooked. While arthroscopic procedures for degenerative knees (including partial meniscectomy and debridement) may be associated with short-term relief, the benefits are limited over time and are associated with a significant rate of adverse events (32). The American Academy of Orthopedic Surgeons, or AAOS, has published an appropriate use criteria guideline regarding the nonarthroplastic management of knee OA, including partial meniscectomy and other arthroscopic procedures, as its treatment options (33). The AAOS has recommended conservative treatment as the initial option of choice. However, according to the AAOS, in select patients with mechanical symptoms (including locking and buckling), partial meniscectomy may be an appropriate practice. Given our results, in addition to available supportive data in the literature opposed to arthroscopic partial meniscectomy (27,28,34), the AAOS may further restrict the indications for partial meniscectomy in such patients. In the follow-up patients in our study, those with mild baseline radiographic features of OA (KL grade, 2) demonstrated higher adjusted hazard for future KR associated with meniscal surgery without a preceding knee trauma when compared with patients with established OA (radiographic KL grade 2). This may be explained by highlighting the point that many patients with established OA (KL grade 2) have already lost a significant amount of their tibiofemoral cartilage (and joint space width) and thus have a lower potential for losing the remaining joint space width. In other words, OA that has already progressed has minimal joint space width left to lose during subsequent follow-up, and JSN progression can best be prevented in subjects with mild OA and preserved joint space widths. These findings, in line with literature reports (35), suggest that in patients with a degenerative meniscal tear in the absence of a preceding trauma and relatively intact cartilage, conservative nonsurgical management may prevent the further risks associated with meniscal surgery (5,36). Other studies have further highlighted the role of the meniscus in the biomechanics and normal function of the tibiofemoral joint. A previous report of the Multicenter Osteoarthritis Study showed that meniscal tears are not the only factors associated with meniscal extrusion (37). Malalignment and accompanying cartilage damage were the other significantly associated factors among subjects with a previous history of knee injury or knee surgery (37). Other studies have confirmed the altered distribution of the pressure load with meniscal damage. In this regard, posterior root tears were suggested to have a pseudo-meniscectomy-like effect by having a peak contact pressure similar to that of total meniscectomy (38). Moreover, even asymptomatic linear intrameniscal signal intensity on MR images was suggested to be a risk factor for further degenerative changes in the knee (39). Ultimately, it is the overall morphologic appearance of the meniscus that is an independent predecessor to knee OA progression (40) and may be used to predict the future need for KR when assessed by using semiquantitative MR imaging scoring systems (21,41). T2 mapping investigations have shown that meniscal tear morphologic appearance and meniscal signal intensity changes are associated with adjacent tibial cartilage damage (42). On the basis of morphologic appearance (42), both traumatic and degenerative meniscal tears may be associated with knee instability, especially in weight-bearing positions (43). Instability reduces the ability of the meniscus to distribute the weight load, which subsequently causes cartilage damage (43). Our study has several limitations. First, the data provided in the OAI cohort were based on patient reports and are therefore amenable to recall and self-reporting bias. We did not perform the analysis on the basis of the surgeons confirmations during meniscal surgery. Second, owing to unavailability of further data stratification in the OAI study, there was no information on the nature of the trauma or the time interval 814 radiology.rsna.org n Radiology: Volume 282: Number 3 March 2017

9 between the preceding trauma, meniscal surgery, and cohort baseline evaluation. Given the severity of OA-related radiographic features, symptoms, and limited physical activity in patients with meniscal surgery (both those with and those without a history of preceding trauma), such data could have been important in delineating the association between preoperative and postoperative time intervals and JSN progression. Time interval data could be helpful when investigating the correlates of worse baseline radiographic OA features in subjects who underwent meniscal surgery. Also, no stratified data were available for the meniscal tear types, side of the tear (medial vs lateral meniscus), and type of surgery performed (complete vs partial meniscectomy), all of which could have affected outcomes. Most of the subjects are assumed to have undergone partial meniscectomy, and if we assume that subjects who have undergone complete meniscectomy are at higher risk for OA progression when compared with those who underwent partial meniscectomy (6), the adjusted HR for subjects who underwent complete meniscectomy may be even higher (20,24). Third, the remainder of the cohort group was a heterogeneous group, with no information on the status of the meniscus at baseline, as some of them might have had a preceding trauma or meniscal tears and received conservative management. Therefore, inclusion of only subjects with a normal meniscus may result in different HR values for patients. Fourth, the number of meniscal surgeries was not included in the analysis, as some patients reported multiple prior meniscus surgeries both with and without preceding knee trauma, and such data are not available in the OAI database. We included them in the meniscal surgery with a preceding knee injury group since, by definition, they had a preceding trauma in their history. Moreover, we could not directly compare the time after meniscal surgery and the time to onset of progression among the study groups. We analyzed the data at the level of individuals (rather than knees, including the worse WOMAC total score or KL grades) and thus could have potentially 11. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relebiased our assessments in association with meniscal surgery. Further clinical trials may elucidate the shortcomings of our study and the timing of JSN progression after meniscal surgery. Finally, we used OARSI scoring to evaluate JSN progression while adjusting our models for baseline KL grades. In other studies, investigators may use KL grade or joint space width (minimum and fixed measurements) (44 46) changes as a measure of OA progression. In fact, using OARSI scoring of JSN allowed us to account for within-grade (as well as fullgrade) progressions of knee OA. In conclusion, in the presence of a reported history of preceding knee trauma and a traumatic onset of symptoms, meniscal surgery was not associated with increased risk of radiographic progression of JSN and KR when adjusted for other known risk factors of OA progression. However, in the absence of such preceding knee trauma, meniscal surgery is associated with increased risk of radiographic JSN features and KR, especially in patients with milder baseline radiographic features of OA (radiographic KL grade, 2). Further investigation in randomized clinical trials is needed to enhance our understanding of the OA outcome in subjects undergoing meniscal surgeries. Acknowledgments: The OAI is a public-private partnership composed of five contracts (N01- AR , N01-AR , N01-AR , N01-AR , and N01-AR ) funded by the National Institutes of Health, a branch of the Department of Health and Human Services, and conducted by the OAI study investigators. Private funding partners include Merck Research Laboratories, Novartis Pharmaceuticals, GlaxoSmithKline, and Pfizer. Private-sector funding for the OAI is managed by the Foundation for the National Institutes of Health. This article was prepared by using an OAI public use data set and does not necessarily reflect the opinions or views of the OAI investigators, the National Institutes of Health, or the private funding partners. Disclosures of Conflicts of Interest: B.Z. disclosed no relevant relationships. N.H.N. disclosed no relevant relationships. F.W.R. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author received money from Boston Imaging Core Lab as chief executive officer and shareholder. Other relationships: disclosed no relevant relationships. A.G. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author received money from MerckSerono, AstraZeneca, Genzyme, OrthoTrophix, and TissueGene for consulting; author received money from Boston Imaging Core Lab as president and shareholder. Other relationships: disclosed no relevant relationships. S.D. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author received money from Toshiba Medical Systems for consulting; author received grants from GERRAF and Carestream Health for a cone-beam computed tomographic clinical trial. Other relationships: disclosed no relevant relationships. References 1. Gu YL, Wang YB. Treatment of meniscal injury: a current concept review. Chin J Traumatol 2010;13(6): Zedde P, Mela F, Del Prete F, Masia F, Manunta AF. Meniscal injuries in basketball players. Joints 2015;2(4): Boutin RD, Fritz RC, Marder RA. Magnetic resonance imaging of the postoperative meniscus: resection, repair, and replacement. Magn Reson Imaging Clin N Am 2014;22(4): Cullen KA, Hall MJ, Golosinskiy A. Ambulatory surgery in the United States, Natl Health Stat Rep 2009 Jan 28;(11): Brophy RH, Matava MJ. Surgical options for meniscal replacement. J Am Acad Orthop Surg 2012;20(5): Hede A, Larsen E, Sandberg H. The long term outcome of open total and partial meniscectomy related to the quantity and site of the meniscus removed. Int Orthop 1992;16(2): Haemer JM, Song Y, Carter DR, Giori NJ. Changes in articular cartilage mechanics with meniscectomy: a novel image-based modeling approach and comparison to patterns of OA. J Biomech 2011;44(12): Pengas IP, Assiotis A, Nash W, Hatcher J, Banks J, McNicholas MJ. Total meniscectomy in adolescents: a 40-year follow-up. J Bone Joint Surg Br 2012;94(12): Petty CA, Lubowitz JH. Does arthroscopic partial meniscectomy always cause arthritis? Sports Med Arthrosc Rev 2012;20(2): Washburn RA, Smith KW, Jette AM, Janney CA. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol 1993;46(2): Radiology: Volume 282: Number 3 March 2017 n radiology.rsna.org 815

10 vant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol 1988;15(12): Hafezi-Nejad N, Guermazi A, Roemer FW, Eng J, Zikria B, Demehri S. Long term use of analgesics and risk of osteoarthritis progressions and knee replacement: propensity score matched cohort analysis of data from the Osteoarthritis Initiative. Osteoarthritis Cartilage 2016;24(4): Kerkhof HJ, Bierma-Zeinstra SM, Arden NK, et al. Prediction model for knee osteoarthritis incidence, including clinical, genetic and biochemical risk factors. Ann Rheum Dis 2014;73(12): Barlow T, Dunbar M, Sprowson A, Parsons N, Griffin D. Development of an outcome prediction tool for patients considering a total knee replacement the Knee Outcome Prediction Study (KOPS). BMC Musculoskelet Disord 2014;15: Pelletier JP, Cooper C, Peterfy C, et al. What is the predictive value of MRI for the occurrence of knee replacement surgery in knee osteoarthritis? Ann Rheum Dis 2013;72(10): Hafezi-Nejad N, Zikria B, Eng J, Carrino JA, Demehri S. Predictive value of semi-quantitative MRI-based scoring systems for future knee replacement: data from the osteoarthritis initiative. Skeletal Radiol 2015;44(11): Altman RD, Fries JF, Bloch DA, et al. Radiographic assessment of progression in osteoarthritis. Arthritis Rheum 1987;30(11): Guermazi A, Hunter DJ, Li L, et al. Different thresholds for detecting osteophytes and joint space narrowing exist between the site investigators and the centralized reader in a multicenter knee osteoarthritis study data from the Osteoarthritis Initiative. Skeletal Radiol 2012;41(2): Sihvonen R, Paavola M, Malmivaara A, et al. Arthroscopic partial meniscectomy versus sham surgery for a degenerative meniscal tear. N Engl J Med 2013;369(26): Stone KR. Current and future directions for meniscus repair and replacement. Clin Orthop Relat Res 1999;(367 Suppl)S273 S Roemer FW, Kwoh CK, Hannon MJ, et al. Can structural joint damage measured with MR imaging be used to predict knee replacement in the following year? Radiology 2015;274(3): Rath E, Richmond JC. The menisci: basic science and advances in treatment. Br J Sports Med 2000;34(4): Messner K, Gao J. The menisci of the knee joint. Anatomical and functional characteristics, and a rationale for clinical treatment. J Anat 1998;193(Pt 2): Choi JY, Chang EY, Cunha GM, Tafur M, Statum S, Chung CB. Posterior medial meniscus root ligament lesions: MRI classification and associated findings. AJR Am J Roentgenol 2014;203(6): Gopez AG, Kavanagh EC. MR imaging of the postoperative meniscus: repair, resection, and replacement. Semin Musculoskelet Radiol 2006;10(3): Vaziri A, Nayeb-Hashemi H, Singh A, Tafti BA. Influence of meniscectomy and meniscus replacement on the stress distribution in human knee joint. Ann Biomed Eng 2008; 36(8): Kijowski R, Woods MA, McGuine TA, Wilson JJ, Graf BK, De Smet AA. Arthroscopic partial meniscectomy: MR imaging for prediction of outcome in middle-aged and elderly patients. Radiology 2011;259(1): Katz JN, Brophy RH, Chaisson CE, et al. Surgery versus physical therapy for a meniscal tear and osteoarthritis. N Engl J Med 2013;368(18): Sihvonen R, Englund M, Turkiewicz A, Järvinen TL; Finnish Degenerative Meniscal Lesion Study Group. Mechanical symptoms and arthroscopic partial meniscectomy in patients with degenerative meniscus tear: a secondary analysis of a randomized trial. Ann Intern Med 2016;164(7): Kim S, Bosque J, Meehan JP, Jamali A, Marder R. Increase in outpatient knee arthroscopy in the United States: a comparison of National Surveys of Ambulatory Surgery, 1996 and J Bone Joint Surg Am 2011;93(11): Salzler MJ, Lin A, Miller CD, Herold S, Irrgang JJ, Harner CD. Complications after arthroscopic knee surgery. Am J Sports Med 2014;42(2): Thorlund JB, Juhl CB, Roos EM, Lohmander LS. Arthroscopic surgery for degenerative knee: systematic review and meta-analysis of benefits and harms. Br J Sports Med 2015;49(19): Sanders JO, Murray J, Gross L. Non-arthroplasty treatment of osteoarthritis of the knee. J Am Acad Orthop Surg 2014;22(4): Kirkley A, Birmingham TB, Litchfield RB, et al. A randomized trial of arthroscopic surgery for osteoarthritis of the knee. N Engl J Med 2008;359(11): Pearse EO, Craig DM. Partial meniscectomy in the presence of severe osteoarthritis does not hasten the symptomatic progression of osteoarthritis. Arthroscopy 2003;19(9): Fetzer GB, Spindler KP, Amendola A, et al. Potential market for new meniscus repair strategies: evaluation of the MOON cohort. J Knee Surg 2009;22(3): Crema MD, Roemer FW, Felson DT, et al. Factors associated with meniscal extrusion in knees with or at risk for osteoarthritis: the Multicenter Osteoarthritis Study. Radiology 2012;264(2): Guermazi A, Hayashi D, Jarraya M, et al. Medial posterior meniscal root tears are associated with development or worsening of medial tibiofemoral cartilage damage: the Multicenter Osteoarthritis Study. Radiology 2013;268(3): Kumm J, Roemer FW, Guermazi A, Turkiewicz A, Englund M. Natural history of intrameniscal signal intensity on knee MR images: six years of data from the Osteoarthritis Initiative. Radiology 2016;278(1): Collins JE, Losina E, Nevitt MC, et al. Semiquantitative imaging biomarkers of knee osteoarthritis progression: data from the FNIH OA Biomarkers Consortium. Arthritis Rheumatol 2016 Apr 25. [Epub ahead of print] 41. Demehri S, Hafezi-Nejad N, Carrino JA. Conventional and novel imaging modalities in osteoarthritis: current state of the evidence. Curr Opin Rheumatol 2015;27(3): Kai B, Mann SA, King C, Forster BB. Integrity of articular cartilage on T2 mapping associated with meniscal signal change. Eur J Radiol 2011;79(3): Barile A, Conti L, Lanni G, Calvisi V, Masciocchi C. Evaluation of medial meniscus tears and meniscal stability: weight-bearing MRI vs arthroscopy. Eur J Radiol 2013;82(4): Oak SR, Ghodadra A, Winalski CS, Miniaci A, Jones MH. Radiographic joint space width is correlated with 4-year clinical outcomes in patients with knee osteoarthritis: data from the Osteoarthritis Initiative. Osteoarthritis Cartilage 2013;21(9): Duryea J, Neumann G, Niu J, et al. Comparison of radiographic joint space width with magnetic resonance imaging cartilage morphometry: analysis of longitudinal data from the Osteoarthritis Initiative. Arthritis Care Res (Hoboken) 2010;62(7): Neumann G, Hunter D, Nevitt M, et al. Location specific radiographic joint space width for osteoarthritis progression. Osteoarthritis Cartilage 2009;17(6): radiology.rsna.org n Radiology: Volume 282: Number 3 March 2017