Nonanatomic femoral tunnel placement increases the risk of subsequent meniscal surgery after ACLR: Part II—Patients without recurrent ACL injury

Abstract

Purpose

The purpose of this study was to identify risk factors for subsequent meniscal surgery following anterior cruciate ligament (ACL) reconstruction (ACLR) in patients without recurrent ACL injury.

Methods

Patients aged ≥14 years who underwent primary ACLR with minimum 1-year follow-up and without recurrent ACL injury were retrospectively reviewed. Patient demographics and surgical data at the time of ACLR were collected. Postoperative radiographs were used to measure femoral and tibial tunnel position, and posterior tibial slope. Univariate and multivariate analyses were performed to identify risk factors for subsequent meniscal surgery.

Results

Of 629 ACLRs that fulfilled the inclusion criteria, subsequent meniscal surgery was performed in 65 [10.3%] patients. Multivariate analysis revealed that medial meniscal repair at the time of ACLR, younger age, anterior femoral tunnel position and distal femoral tunnel position were significantly associated with subsequent meniscal surgery (p < 0.001, p = 0.016, p = 0.015, p = 0.035, respectively). The frequency of femoral tunnel placement >10% outside of the literature-established anatomic position was significantly higher in those who underwent subsequent meniscal surgery compared to those who did not (38.3% vs. 20.3%, p = 0.006). Posterior tibial slope and ACL graft type were not significantly associated with subsequent meniscal surgery.

Conclusion

Medial meniscal repair at the time of ACLR, younger age and nonanatomic femoral tunnel placement were risk factors for subsequent meniscal surgery in patients without recurrent ACL injury. Femoral tunnel placement <10% outside of the native anatomic position is important to reduce the risk of subsequent meniscal surgery.

Level of Evidence

Level IV.

Abbreviations

ACL

anterior cruciate ligament

ACLR

anterior cruciate ligament reconstruction

Allo

allograft

A-P

anterior–posterior

BMI

body mass index

BPTB

bone-patellar tendon-bone

HS

hamstring tendon

ICC

intraclass correlation coefficient

P-D

proximal–distal

QT

quadriceps tendon

INTRODUCTION

Subsequent meniscal surgery after anterior cruciate ligament (ACL) reconstruction (ACLR) is a concern for patients who undergo primary ACLR. A previous study following 3276 patients who underwent ACLR found that 11.9% underwent a subsequent surgery for meniscus-related pathology after 6 years of follow-up, which was one of the most common subsequent procedures [27]. Previously described risk factors for subsequent meniscal surgery in patients undergoing ACLR include medial meniscus repair at the time of ACLR, young age and high activity level [9, 17, 24, 27]. Most of these studies included meniscus surgeries performed concomitantly with a subsequent revision ACLR and meniscus injuries may have resulted from ACL graft retear or failure. Therefore, risk factors for additional meniscus surgery in the absence of recurrent ACL injury are not clear.

Appropriately placing the ACL graft within the native insertion sites during primary ACLR is important to protect the meniscus from further injury since the ACL and menisci are biomechanically interdependent [22]. A previous study (Part I) from our institution described an increased risk of subsequent meniscal surgery for patients with an anterior femoral tunnel position, but included patients who experienced a recurrent ACL tear [14]. In addition, the Part I study did not account for posterior tibial slope or preoperative patient activity, which may be additional risk factors for meniscus injury [4]. Pre- and postoperative radiographs are useful to measure posterior slope, tunnel position and static anterior tibial translation. Practice patterns have recently emphasised taking adequate strict lateral postoperative radiographs more frequently for these reasons, which may also strengthen conclusions in the current study.

The purpose of this study was to determine risk factors for subsequent meniscal surgery following primary ACLR in patients without recurrent ACL injury. Nonanatomic tunnel position of the ACL graft and increased posterior tibial slope were hypothesised to be risk factors, along with medial meniscus repair at the time of ACLR, younger age and higher activity level.

MATERIALS AND METHODS

Institutional Review Board approval was obtained to conduct this retrospective review (STUDY19100047). Informed consent for the study was not required. Patients aged ≥14 years who underwent primary ACLR at our institution from 2014 to 2022 carried out by seven high-volume sports fellowship-trained orthopaedic surgeons were included. Additionally, a minimum follow-up of at least 1 year was required for study inclusion. Exclusion criteria were patients with (1) multiligament knee surgery; (2) concomitant osteotomy, lateral extra-articular tenodesis or meniscus allograft (Allo) transplantation; (3) double-bundle or over-the-top ACLR; (4) revision ACLR; (5) recurrent ACL injury defined as revision ACLR, magnetic resonance imaging-confirmed ACL tear or residual anterior instability with a soft endpoint and asymmetry on Lachman test and (6) any history of prior surgery on the ipsilateral knee.

Demographic data, including patient age, sex and body mass index, along with other variables including Marx Activity Rating Scale prior to injury and time from injury to ACLR, were collected from medical records. Time from injury to ACLR was classified into four groups (<3 weeks, 3 weeks–3 months, 3–9 and >9 months) as described previously [10]. Operative characteristics, including concomitant meniscal surgery, meniscus tear type both at the time of ACLR and subsequent surgery and ACL graft type (hamstring tendon, quadriceps tendon, bone-patellar tendon-bone and Allo), were obtained from operative reports. The primary outcome was any subsequent reoperation for meniscus-related pathology as evaluated clinically by the operating surgeon. The femoral tunnel was placed through the anteromedial portal with a flexible pin and reamer.

Radiographic measurements

Postoperative strict lateral knee radiographs were used to assess the position of ACL graft tunnels and posterior tibial slope based on previously described methods [2, 15, 21]. All radiographic measurements were performed by one of three observers. The measurements were only performed if a strict lateral image was available, defined as an offset of less than 6 mm between the posterior halves of the medial and lateral femoral condyles, as this has been shown to maximise inter-observer reliability [23]. The centre of the femoral tunnel was reported as a percentage of the anterior–posterior (A-P) and proximal–distal (P-D) direction corresponding to the A-P dimension of the lateral femoral condyle along Blumensaat's line and the perpendicular P-D line (Figure 1a) [15]. The centre of the tibial tunnel was reported as a percentage of the A-P width of the tibia [15]. Each tunnel position was categorised according to upper and/or lower quartiles: ≤30% and >30% in the A-P direction for the femoral tunnel, <28%, 28%–39% and >39% in the P-D direction for the femoral tunnel and <35% and ≥35% in the tibial tunnel.

Figure 1

Radiographic measurements for tunnel position and posterior tibial slope. (a) Femoral tunnel position was described as a percentage of the anterior–posterior (A-P) direction (line 3/line 1) and the proximal–distal (P-D) direction (line 4/line 2). Tibial tunnel position was measured as a percentage of the anterior–posterior direction (line 6/line 5). (b) The distance from reported anatomic position [6] (white circle: A-P direction; 25%, P-D direction; 29%) to the centre of the femoral tunnel position (black circle) was calculated and standardised by dividing by the condylar width (line 1). The yellow dashed circle indicates the outline of the femoral tunnel. (c) Posterior tibial slope was measured as the angle between the tibial plateau and the line tangent to the anatomic axis of the tibia using the ‘the circle three-point method’ [2, 21]. (d) Anatomic tunnel placement was determined by ≤10% difference from native insertion. (d) Anatomic tunnel placement, while (b) shows nonanatomic tunnel placement (>10% difference from native insertion).

In addition, the distance from the centre of the femoral tunnel to the anatomic centre of the femoral ACL footprint was calculated based on a previous study describing the anatomic centre of the native femoral ACL footprint (A-P direction; 25%, P-D direction; 29%) [6]. The difference in distance between the femoral tunnel and the anatomic ACL footprint in both dimensions was standardised to the A-P length of the lateral femoral condyle (Figure 1a; line 1) and reported as a percentage. Additionally, the mean A-P length of the lateral femoral condyle was approximately 5 cm, making a 10% difference in tunnel position equivalent to a difference of 5 mm from the anatomic insertion. Using this threshold, femoral tunnel position was categorised into two groups corresponding to anatomic (≤10% difference from native insertion) placement and nonanatomic (>10% difference from native insertion) placement (Figure 1b,d). Patients were classified into anatomic and nonanatomic groups based on the measured distance between the anatomic insertion and the femoral tunnel relative to the A-P length of the lateral femoral condyle. Any distance greater than 10% of the condylar length, rather than only one of the A-P or P-D dimension distances, was considered nonanatomic.

Measurements for posterior tibial slope required at least 10 cm of the tibial shaft cortices visible. Posterior tibial slope was measured as the angle between the line tangent to the anatomic axis of the tibia and superimposed cortical densities of the medial and lateral tibial plateau using ‘the circle 3-point method’ (Figure 1c) [2, 21]. Posterior tibial slope was classified into two groups, ≤12° and >12°, as 12° has been used as a cutoff value for an increased slope [21].

Statistical analysis

To calculate intra- and interobserver reliability for the measurements for the femoral and tibial tunnel position and posterior tibial slope using plain radiographs, measurements were independently performed by three observers and repeated one month after the first measurement by one observer for 50 randomly selected patients. Intra- and interobserver reliability was assessed by the intraclass correlation coefficient (ICC_1,1 and ICC_2,1, respectively). ICC was interpreted as poor (<0.50), moderate (0.50–0.75), good (0.75–0.90) or excellent (>0.90) [18].

Continuous variables were reported as mean ± SD, while dichotomous variables were reported as the number and percentage. Patients were classified into two groups: subsequent meniscal surgery and no subsequent meniscal surgery. For patients undergoing subsequent meniscal surgery, meniscus tear treatment (repair or meniscectomy) at the time of additional surgery was compared to the meniscus tear treatment at the time of initial ACLR. To compare the variables between groups, a χ² test was used for categorical variables. An independent-samples t test or the Mann–Whitney U test was used for continuous variables depending on the normality as assessed by the Shapiro–Wilk test. In addition, variables associated with subsequent meniscal surgery (p < 0.200) were entered into a multivariable model using manual backwards stepwise binary logistic regression analysis (p > 0.100 removed). Only cases with no missing data for all variables entered into the multivariate analysis were included. All statistical analyses were performed using SPSS software version 28.0.1.1 (IBM). Statistical significance was set at p < 0.050.

The post hoc power analysis was performed using the χ² goodness-of-fit test with an α error of 0.05 and an effect size of 0.2 to detect the difference between groups, which showed that the power in the study was >0.9. Statistical power was calculated using G* Power v3.1.9 (Heinrich Heine University).

RESULTS

Out of 2079 primary ACLRs identified, 629 ACLR fulfilled the inclusion criteria (mean age, 24.8 ± 10.0 years; 310 female [49.3%]; mean follow-up duration, 2.3 ± 1.6 years; Figure 2). Mean intra- and interobserver reliability showed good or excellent reliability in all measurements (Table 1). Subsequent meniscal surgery was performed in 65/629 (10.3%) patients (medial, 45/65 [69.2%]; lateral, 15/65 [23.1%], both 5/65 [7.7%]). Patients with a meniscus repair at the time of ACLR were more likely to undergo subsequent meniscectomy compared to those who initially had a meniscus tear treated with meniscectomy or no surgery at the time of ACLR (p = 0.015; Table 2).

Figure 2

Flow diagram for enrolment. ACLR, anterior cruciate ligament reconstruction; n, number of patients.

Table 1. Intra- and interobserver reliability for radiographic measurements.

	Femoral tunnel A-P	Femoral tunnel P-D	Tibial tunnel	Posterior tibial slope
ICC1,1	0.847 (0.746–0.910)	0.751 (0.600–0.850)	0.782 (0.647–0.870)	0.797 (0.669–0.879)
ICC2,1	0.902 (0.772–0.961)	0.813 (0.575–0.927)	0.816 (0.560–0.935)	0.801 (0.646–0.897)

• Note: Data are reported as mean (95% CI). ICC, interclass correlation coefficient. A-P, anterior–posterior. P-D, proximal–distal.
• Abbreviations: A-P, anterior–posterior; CI, confidence interval; ICC, interclass correlation coefficient; P-D, proximal–distal.

Table 2. Comparison of meniscal tear treatment at the time of additional surgery and initial ACLR.

	Meniscus repair at the time of additional surgery	Meniscectomy at the time of additional surgery	p Value
Meniscus repair at the time of ACLR (n = 48)	7/48 (14.6%)	41/48 (85.4%)
Meniscectomy or left in-situ at the time of ACLR (n = 22)	9/22 (40.9%)	13/22 (59.1%)
			0.015

• Note: Data are reported as count (%); statistical significance (p < 0.05) is indicated in bold.
• Abbreviation: ACLR, anterior cruciate ligament reconstruction.

The patients were significantly younger in the subsequent meniscal surgery group than in the no subsequent meniscal surgery group (22.1 ± 8.0 vs. 25.1 ± 10.2 years, p = 0.044). Moreover, the rate of medial meniscus repair at the time of ACLR was significantly greater in the subsequent meniscal surgery group than in the no subsequent meniscal surgery group (42/65 [64.6%] vs. 145/564 [25.7%], p < 0.001; Table 3). Meniscus tear type or repair method was not different between groups (Table A1).

Table 3. Comparison of patient characteristics between groups.

Variable	Subsequent meniscal surgery (n = 65)	No subsequent meniscal surgery (n = 564)	p Value
Age (years)	22.1 ± 8.0	25.1 ± 10.2	0.044
Sex (female)	34 (52.3%)	276 (48.9%)	n.s.
Body mass index (kg/m^2)	26.2 ± 4.8	26.8 ± 5.6	n.s.
ACL graft type [HS/QT/BPTB/Allo]	11/18/20/16	127/157/151/129	n.s.
Medial meniscus repair at the time of ACLR	42 (64.6%)	145 (25.7%)	<0.001
Lateral meniscus repair at the time of ACLR	16 (24.6%)	114 (20.2%)	n.s.
Medial meniscectomy at the time of ACLR	3 (4.6%)	52 (9.2%)	n.s.
Lateral meniscectomy at the time of ACLR	10 (15.4%)	64 (11.3%)	n.s.

• Note: Data are reported as mean ± SD in the case of continuous variables or count (%) in the case of categorical variables. Statistical significance (p < 0.050) is indicated in bold.
• Abbreviations: ACLR, anterior cruciate ligament reconstruction; Allo, allograft; BPTB, bone-patellar tendon-bone; HS, hamstring tendon; QT, quadriceps tendon.

Time from injury to ACLR was comparable between groups. Of the 209 patients with available baseline Marx Activity Rating Scale scores, patients in the subsequent meniscal surgery group had a significantly higher score than those in the no subsequent meniscal surgery group (13.9 ± 4.3 vs. 12.4 ± 4.7, p = 0.013; Table 4).

Table 4. Comparison of surgical timing and baseline activity between groups.

Variable	Subsequent meniscal surgery	No subsequent meniscal surgery	p Value
Time from injury to primary surgery	n = 64	n = 557	n.s.
<3 weeks	19 (29.7%)	111 (19.9%)
3 weeks–3 months	31 (48.4%)	327 (58.7%)
3–9 months	10 (15.6%)	77 (13.8%)
>9 months	4 (6.3%)	42 (7.5%)
Baseline Marx Activity Rating Scale	13.9 ± 4.3 (n = 24)	12.4 ± 4.7 (n = 185)	0.013

• Note: Data are reported as mean ± SD in the case of continuous variables or counts (%) in the case of categorical variables. Statistical significance (p < 0.050) is indicated in bold.

Of the 629 included patients, 401 had strict lateral radiographic films available for appropriate measurements. One patient was excluded from the measurement for tibial tunnel position since the tibial tunnel was not visible. The femoral tunnel position was significantly more anterior in the subsequent meniscal surgery group compared to the no subsequent meniscal surgery group (p = 0.024). In addition, a significant difference was found between groups for femoral tunnel position in the P-D direction (p = 0.010). Residual analysis showed that the rate of distal femoral tunnel position was greater in the subsequent meniscal surgery group than the no subsequent meniscal surgery group (p < 0.050). Furthermore, femoral tunnel placement >10% outside of the anatomic position was found more frequently in the subsequent meniscal surgery group (p = 0.006). No significant difference was detected between groups with respect to tibial tunnel position and posterior tibial slope (Table 5).

Table 5. Comparison of tunnel position and tibial posterior slope between groups.

Variable	Subsequent meniscal surgery	No subsequent meniscal surgery	p Value
Femoral tunnel A-P	n = 47	n = 354	0.024
(>30%, anterior)	18 (38.3%)	82 (23.2%)
Femoral tunnel P-D	n = 47	n = 354	0.010
(<28%, proximal)	9 (19.1%)	94 (26.6%)
(28%–39%)	18* (38.3%)	181* (51.1%)
(>39%, distal)	20* (42.6%)	79* (22.3%)
Femoral tunnel difference from anatomic position	n = 47	n = 354	0.006
(>10%, nonanatomic)	18 (38.3%)	72 (20.3%)
Tibial tunnel A-P	n = 47	n = 354	n.s.
(<35%, anterior)	7 (14.9%)	84 (23.7%)
Tibial posterior slope	n = 44	n = 303	n.s.
(>12°)	9 (20.5%)	65 (21.5%)

• Note: Data are reported as count (%); the statistical significance (p < 0.050) is shown in bold.
• Abbreviations: A-P, anterior–posterior, P-D, proximal–distal.
• * p < 0.050 (residual analysis).

The Marx Activity Rating Scale was excluded from the multivariate logistic among variables with p < 0.200 in the univariate analysis due to a large number of missing data. Considering collinearity, femoral tunnel difference from the anatomic position and the femoral tunnel position in the A-P and P-D dimensions was entered separately into the multivariate logistic analysis. In a multivariate logistic analysis of 401 patients (subsequent meniscal surgery group; n = 47/401 [11.7%], no subsequent meniscal surgery group; n = 354/401 [88.3%]), age, medial meniscus repair, a more anterior femoral tunnel A-P distance and a more distal femoral tunnel P-D distance were found to be significantly associated with subsequent meniscal surgery (p = 0.016, p < 0.001, p = 0.015, p = 0.035, respectively; Table 6). The multivariate analysis including age, medial meniscus repair, femoral tunnel difference from anatomic position and tibial tunnel position showed that age, medial meniscus repair and femoral tunnel difference from anatomic position were significantly associated with subsequent meniscal surgery (p = 0.015, p < 0.001, p = 0.002, respectively; Table 6).

Table 6. Multivariate logistic analysis of risk factors for subsequent meniscal surgery including A-P and P-D femoral tunnel position or femoral tunnel difference from anatomic position.

Variable	B	OR	95% CI	p Value
Variable including A-P and P-D femoral tunnel position
Age (year)	−0.053	0.948	0.909–0.990	0.016
Medial meniscus repair	1.383	3.988	2.084–7.632	<0.001
Femoral tunnel A-P	0.843	2.322	1.180–4.568	0.015
Femoral tunnel P-D	0.500	1.648	1.036–2.621	0.035
Variable including femoral tunnel difference from anatomic position
Age (year)	−0.052	0.949	0.910–0.990	0.015
Medial meniscus repair	1.467	4.334	2.256–8.324	<0.001
Femoral tunnel difference from anatomic position	1.055	2.872	1.451–5.684	0.002

• Note: Statistical significance (p < 0.050) is indicated in bold. The analysis was performed with 401 cases, excluding cases with any missing data for the variables.
• Abbreviations: A-P, anterior–posterior; CI, confidence interval; OR, odds ratio; P-D, proximal–distal.

DISCUSSION

The most important finding of this study was that femoral tunnel placement >10% outside of the native anatomic position was a significant risk factor for subsequent meniscal surgery after primary ACLR in patients without a recurrent ACL injury. In addition, other risk factors for subsequent meniscal surgery included (1) medial meniscus repair at the time of ACLR, (2) younger age and (3) higher baseline Marx Activity Rating Scale. Contrary to our hypothesis, posterior tibial slope and tibial tunnel position were not significantly associated with subsequent meniscal surgery. Furthermore, ACL graft type was not associated with the need for additional surgery.

Numerous studies have demonstrated the importance of anatomic ACL graft tunnel position during ACLR [7, 8, 20]. A cadaveric study showed that an anatomic femoral tunnel position restored knee kinematics closer to the ACL intact knee compared to the nonanatomic position [20]. Nonanatomic femoral tunnel placement has also been shown to increase the risk of revision ACLR, cartilage injury and bucket-handle meniscus injury [7, 8]. Consistent with previous findings [8, 14], the present study found that nonanatomic femoral tunnel positions, specifically an anteriorly placed tunnel, were significant risk factors for subsequent meniscal surgery. However, the present findings also demonstrated distal femoral tunnel position to be a risk factor, contrary to results of the Part I study showing no association [14]. In addition, an assessment of the relative difference between femoral tunnel position and the native femoral ACL footprint also showed that nonanatomic femoral tunnel position (>10% difference) was considerably higher in the subsequent meniscal surgery group (38%) compared to the no surgery group (20%). Since tunnel drilling is dependent on surgeon technique, surgeons should remain mindful to create the tunnel in the native anatomic position to reduce the risk of subsequent meniscal surgery.

Meanwhile, tibial tunnel placement was not a significant risk factor for subsequent meniscal surgery. A study investigating the outcome 10–12 years after ACLR demonstrated that a posteriorly placed tibial tunnel was associated with increased rotatory knee instability [16]. Furthermore, a cadaveric study found that tibial tunnel position in the posterolateral portion of the tibial footprint resulted in more rotational laxity than tunnel placement in the anteromedial portion of the tibial footprint regardless of the femoral tunnel position [30]. Variations in tibial tunnel placement in this study may not have been large enough to generate rotational instability that would affect additional meniscus surgery.

In the present study, the posterior tibial slope was not associated with subsequent meniscal surgery. An increased posterior tibial slope has also been reported to increase the risk of ACL injury and graft failure [26, 29]. However, evidence discussing the association between posterior tibial slope and meniscus injury is controversial. Some studies show that meniscus root tears are increased with steeper posterior tibial slopes [11, 32], while another study reported that a smaller posterior tibial slope was associated with meniscal tears requiring surgery in ACL intact knees [4]. Our study demonstrated that a posterior tibial slope was not associated with the need for additional meniscal surgery in patients who underwent ACLR.

Although a high healing rate of medial meniscus repairs (74%–96%) accompanied by ACLR has been reported [1, 28], the medial meniscus is susceptible to reinjury after ACLR. A study investigating concomitant medial meniscus repairs with ACLR showed that 35% of completely or incompletely healed tears had a new meniscus injury on second-look arthroscopy [28]. A systematic review also demonstrated higher reoperation rates following combined ACLR and medial meniscus repair compared to combined ACLR and meniscectomy (13% vs. 1%), although repair better restored anterior knee stability [25]. Consistent with previous studies [16, 24, 27], medial meniscus repair at the time of ACLR was the risk factor resulting in the highest odds for subsequent meniscal surgery (odds ratio: 4.0–4.3). Subsequent meniscus surgery also involved meniscectomy more commonly than meniscus repair in the present study, which was comparable to rates of a prior study that also showed higher rates of total knee arthroplasty with meniscectomy at a mean follow-up of 18 years [13].

The finding that concomitant medial meniscus repair was a risk factor, whereas lateral meniscus repair was not, is also consistent with previous studies showing that the risk of meniscus repair failure differs between medial and lateral repairs [17, 27, 31]. Medial meniscus repair has been reported to fail earlier than lateral meniscus repair with concomitant ACLR (2.1 vs. 3.7 years) [31]. Follow-up in the present study may, therefore, be limited in its ability to capture all reoperations for lateral meniscus pathology. The lower mobility and increased biomechanical load of the medial meniscus compared to the lateral meniscus could affect the difference in meniscus surgery after ACLR [12]. Furthermore, the medial meniscus provides secondary stabilisation for translational stability as demonstrated in a previous study showing further increased anterior translation up to 58% at 90° of knee flexion following meniscectomy in ACL deficient knees [19]. A biomechanics study also demonstrated that the resultant force in the medial meniscus of ACL deficient knees was significantly larger than that in the ACL intact knee [3]. This may account for the higher frequency of medial meniscus injuries in chronic ACL tears [5] and in post-ACLR knees, where slight instability remains due to nonanatomic tunnel placement. The resultant laxity with nonanatomic tunnel placement is supported by the present findings as nonanatomic tunnels were found in 38% of those who underwent subsequent meniscal surgery compared to 20% in those who did not.

The current study showed that younger age and higher Marx Activity Scoring Scale were risk factors, consistent with previous studies [24, 27]. Although the present study was unable to include Marx scores in the multivariate analysis, higher activity has been reported as a risk factor for subsequent surgeries after ACLR not only for meniscus but also for cartilage-related pathology [24]. However, younger age was significantly associated with a subsequent meniscus surgery on multivariate analysis and young age is often correlated with high activity levels.

The present study had some limitations. First, a mean follow-up of 2.3 years may not fully capture all cases of subsequent meniscal surgery. However, the effect of residual instability due to nonanatomic tunnel position on subsequent meniscal surgery may be observed in a relatively short period of time after return to sports. Furthermore, it was retrospective and subject to the inherent limitations of limited data and follow-up. For example, Marx scores were not able to be included in the multivariate analysis due to excessive missing data and the availability of strict lateral radiographs was limited. In addition, some patients received follow-up outside of our healthcare system, which may underestimate the rate of subsequent meniscal surgery. Moreover, bone tunnel position was measured by radiography and although measurement reliabilities were good, they were not excellent. Computed tomography (CT) would provide a more accurate assessment, although CTs are rarely obtained in the absence of planned revision ACLR. Despite these limitations, the current study is valuable in demonstrating the association between ACL graft femoral tunnel position and subsequent meniscal surgery and underscores the importance of anatomic ACLR even for patients who do not experience a recurrent ACL injury. Surgeons should avoid anterior femoral tunnel placement during ACLR as this may increase the risk of subsequent meniscus surgery in the absence of a recurrent ACL injury.

CONCLUSION

Medial meniscal repair at the time of ACLR, younger age and nonanatomic femoral tunnel placement were risk factors for subsequent meniscal surgery in patients without recurrent ACL injury. Femoral tunnel placement <10% outside of the native anatomic position during initial ACLR is important to reduce the risk of subsequent meniscal surgery.

AUTHOR CONTRIBUTIONS

Literature search, study design and primary manuscript preparation were performed by Jumpei Inoue, Joseph D. Giusto and Jonathan D. Hughes. Data collection, analysis and interpretation of results were performed by Jumpei Inoue, Koji Nukuto, Joseph D. Giusto and Sahil Dadoo. Statistical analysis was performed by Jumpei Inoue and Koji Nukuto. Final manuscript drafting and editing were performed by Jonathan D. Hughes, Bryson P. Lesniak and Volker Musahl. All listed authors have contributed substantially to this work.

ACKNOWLEDGEMENTS

The authors have no funding to report.

CONFLICTS OF INTEREST STATEMENT

V. M. reports educational grants, consulting fees and speaking fees from Smith & Nephew plc and educational grants from Arthrex and DePuy/Synthes, is a board member of the International Society of Arthroscopy, Knee Surgery and Orthopaedic Sports Medicine (ISAKOS) and deputy editor-in-chief of Knee Surgery, Sports Traumatology, Arthroscopy (KSSTA). J. D. H. is on the Editorial Board of Knee Surgery, Sports Traumatology, Arthroscopy (KSSTA).

ETHICS STATEMENT

This study was approved by the Institutional Review Board (IRB), that is, the Ethics Committee of University of Pittsburgh, in Pittsburgh, USA (reference ID no. STUDY19030196).

APPENDIX A

Table A1. Comparison of meniscal tear type and meniscal repair method between groups.

Variable	Subsequent meniscal surgery (n = 65)	No subsequent meniscal surgery (n = 564)	p Value
Medial meniscus tear type at the time of ACLR	n = 45	n = 197	n.s.
Vertical (longitudinal)	35 (77.8%)	112 (56.9%)
Radial	1 (2.2%)	5 (2.5%)
Oblique (parrot beak)	0 (0.0%)	4 (2.0%)
Bucket handle	7 (15.6%)	33 (16.8%)
Horizontal	1 (2.2%)	9 (4.6%)
Complex	1 (2.2%)	26 (13.2%)
Root	0 (0.0%)	4 (2.0%)
Fraying	0 (0.0%)	1 (0.5%)
Unknown	0 (0.0%)	3 (1.5%)
Lateral meniscus tear type at the time of ACLR	n = 26	n = 178	n.s.
Vertical (longitudinal)	5 (19.2%)	64 (36.0%)
Radial	3 (11.5%)	30 (16.9%)
Oblique (parrot beak)	1 (3.8%)	8 (4.5%)
Bucket handle	0 (0.0%)	8 (4.5%)
Horizontal	4 (15.4%)	10 (5.6%)
Complex	6 (23.1%)	24 (13.5%)
Root	7 (26.9%)	28 (15.7%)
Fraying	0 (0.0%)	5 (2.8%)
Unknown	0 (0.0%)	1 (0.6%)
Meniscal repair method at the time of ACLR	n = 58	n = 259	n.s.
All-inside	48 (82.3%)	216 (83.4%)
Inside-out	1 (1.7%)	5 (1.9%)
Out-side-in	0 (0%)	2 (0.7%)
Pull-out (root repair)	6 (10.3%)	25 (11.6%)
Combined	3 (5.2%)	11 (4.2%)

• Abbreviation: ACLR, anterior cruciate ligament reconstruction.

Open Research

DATA AVAILABILITY STATEMENT

Anonymized data from the study are available upon reasonable request.