Evaluation of retear rates following transtibial anterior cruciate ligament reconstruction: Functional outcomes and GNRB assessment with a mean 6.3-year follow-up

Abstract

The method of reconstruction of the anterior cruciate ligament (ACL) using the tibia-dependent femoral tunnel technique is highly criticized. It would not allow anatomical placement of the graft and would therefore lead to a high rate of retear. This retrospective study aimed to evaluate the rate of retear in ACL reconstructions using the transtibial tunnel (TT) technique and assess functional outcomes. The study included 148 patients (153 knees) who underwent ACL reconstruction with a minimum two-year follow-up period. Patients underwent clinical examinations, completed functional questionnaires (KOOS, Lysholm, ACL-RSI), and underwent laximetry measurements with GNRB at 134 N and 200 N. Two patients had retears detected on MRI and underwent revision surgery prior to the study, leaving 151 knees for evaluation. The retear rate was 6.6 %, with an overall rate of 7.8 % when including revision cases. Laxity measurements obtained with GNRB showed a strong correlation with functional scores, indicating worse scores with greater laxity differences (p < 0.0001). The retear rate observed in this study was at the higher end of those reported in the literature. The study indications for TT ACL reconstructions must be defined more specifically to reduce failure rates.

Level of evidence: III (retrospective cohort study).

1. Introduction

Rupture of the anterior cruciate ligament (ACL) is one of the most frequent injuries in sports traumatology. In the United States, ACL tears have been estimated to occur in one out of 3000 individuals, with an average 100,000 surgical reconstructions yearly.^1,2 The benefits of ACL reconstruction surgery have been proven,³ and several repair techniques have been proposed. Nonetheless, graft retears remain the most common cause for failed ACL surgery, with an overall 5.8 % retear rate reported in the literature, irrespective of technique.⁴ The type of graft,⁵ the method graft fixation and placement⁶ and the reconstruction method⁷ have been extensively studied to better understand the causes of retear, but the data remains inconclusive. More recently, the importance of combining reconstruction of the anterolateral ligament (ALL) with that of the ACL using an out-in technique^8,9 has been emphasized, particularly in cases of rotational instability. This approach has been reported to reduce the retear rate to 4.5 %, better stability and faster return to sport.^10,11

Anatomic graft placement, especially at the femoral end, may explain the lower failure rate of tunnel independent technique, using either the anteromedial portal or retrograde drilling. Although not the most anatomic, the TT technique offers other advantages, including being a shorter procedure, less technically demanding, and allowing for tunnel positioning that restores ligamentous isometry.13, 14, 15, 16, 17 Some studies on this technique report comparable or even lower retear rates than those found with the outside-in anteromedial portal technique.¹⁸

Evaluating the efficacity or the retear rate of ligament reconstructions, outside an obvious traumatic injury, can be difficult in practice. Functional scores are commonly used but more subjective,¹⁹ while objective clinical or radiological knee laxity measurements with a dedicated device are rarely chosen.²⁰ Among these, the GNRB (Genourob, Laval, France) is a validated assessment tool,21, 22, 23 with better reliability and reproducibility than other existing methods, such as the KT1000 (Aircast Europa, Neubeuern, Allemagne)²¹ or the Telos (Telos Medical, Millersville, MD, USA).²⁴ Thus far, the outcomes of TT technique has never been objectively evaluated.^19,25

The primary objective of this study was to evaluate the rate of retear of ACL reconstructions using the TT technique with objective laximetric measurement using the GNRB. The secondary objective was to evaluate the functional outcome at the last follow-up. Our hypothesis was that the TT technique would not have a higher retear rate than other ACL reconstructions reported in the literature that are deemed more anatomic.

2. Materials and methods

2.1. Population

This single center retrospective cohort study was performed in a University Hospital. Between January 1st^, 2006 and December 31st^, 2016, 165 patients underwent primary ACL reconstruction with the TT technique. To be included, patients needed a minimum 2-year follow-up. Seventeen patients refused to participate. Thus, 148 patients (107 male/41 female), 5 bilateral cases, were included for a total of 153 knees. Two patients had already undergone revision surgery following retear seen on MRI. The mean age was 35 years (16–53). The mean body mass index (BMI) was 25.7 kg/m² (18–49). One hundred five patients underwent a bone-patellar tendon-bone (BPTB) graft and 48 had a hamstring (HS) graft using the semitendinosus and gracilis tendons (Fig. 1). Fifty-nine patients had a meniscus injury that was discovered intraoperatively; resection was performed for 32, one was repaired and 14 remained untreated (Table 1).

Fig. 1.

Patients flow diagram.

Table 1.

Comparative preoperative data of patients according to the graft type.

	BPTB (n = 105)	Hamstring graft (n = 46)
Sex
Male	91 (87 %)	18 (37.5 %)
Female	14 (13 %)	28 (62.5 %)
Age (years)	30 [15–52]	27 [16–49]
BMI (kg/m2)	26 [18–49]	24 [19–40]
Preoperative Tegner	4.5 [0–7]	4.2 [0–9]
Meniscal lesion	41 (39 %)	18 (37.5 %)

2.2. Surgical technique

The first surgical step was graft harvest and preparation. The second step focused on arthroscopic joint exploration. ACL rupture was confirmed; meniscal and articular cartilage lesions were assessed and repaired when needed.

The third step consisted in bone tunnel preparation. The tibial aiming guide was positioned, with the entry point supported on the anteromedial tibial cortex and the intra-articular exit point at the footprint of the torn ACL. The guide was angled between 45° and 55°. The eyelet pin was pulled, making sure there was no impingement with the notch, before drilling the tibial tunnel according to the diameter of the prepared graft. An offset femoral guide was inserted through the tibial tunnel, with the knee flexed between 70° and 90°. The transtibial femoral tunnel was made, maintaining a distance of approximately 5 mm between the chosen femoral entry point and the posterior femoral cortex of the femoral condyle. The eyelet guide pin was removed, and the femoral socket drilled, based on the length and diameter of the end of the previously sutured graft.

Finally, femoral fixation was undertaken with the RIGIDFIX (DePuy Mitek Inc., Raynham, MA, USA) instrumentation system, followed by graft ascension through the tunnels. This was performed arthroscopically, bringing the graft from the tibia to the femur with a traction suture. Once the graft butted against the femoral socket, graft fixation was completed with two orthogonal RIGIDFIX pins. Tibial fixation in 20° knee flexion was completed. With the BPTB graft, we used an interference screw 1 mm smaller than the diameter of the bone plug, and for the HS graft, a peg mounted with an interference screw of the same diameter as the graft. We confirmed there was no impingement with the notch at the end of the surgery.

The rehabilitation protocol was the same for both graft types. Immediate weight-bearing was allowed with limited 90° knee flexion in a knee brace for 1 month. Thereafter, free range of motion was possible. Return to inline sports and then to pivot sports was allowed at 6 and 9 months, respectively.

2.3. Subjective functional and clinical assessment

All the patients were reviewed in person during the first six months of 2019. They were questioned about return to sports, when and at which level, whether they had complaints about instability and locking, as well as their satisfaction level. They were also asked if they had suffered any new complications or undergone revision surgery (in our hospital or any other) in the operated knee since the last follow-up. Before interviewing the patient, a detailed analysis of the surgical and follow-up records was conducted to collect all the data on the surgical technique used (nature of the graft, tunnel characteristics, fixation method, intraoperative events, associated procedures) as well as postoperative data (type and duration of immobilization, rehabilitation protocol, time to return to sports). Finally, patients were asked if they felt any instability at the final follow-up.

Clinical outcomes consisted in a standardized physical examination, bilateral and comparative, which included ROM, the drawer test, the Lachman test, laxity in the frontal plane, and thigh circumference measured 20 cm from the proximal patellar pole with a sewing tape.

Functional outcomes were assessed with the Tegner score²⁶ at the time of the review and preoperatively. Three other functional scores were also administered at the last follow-up: the Knee Injury and Osteoarthritis Outcome Score (KOOS),²⁷ the Lysholm,²⁸ and the Anterior Cruciate Ligament Return to Sport After Injury (ACL-RSI).²⁹

2.4. Objective knee laxity assessment

We chose the validated GNRB system (Genourob, Laval, France) to quantify knee laxity.^24,31,32 Because this was a new device for our team, the designated examiner underwent several training sessions on its workings. When measuring knee laxity, the patient was half-seated on the examination table, lower limb resting on the device and held by two fixed points at the patellar and ankle level. The sensor was positioned on the tibial tuberosity, two finger widths from the support. A posterior driven force was applied to mimic the anterior drawer. Measurements were taken twice for each patient and for each knee, once at 134 N and then at 200 N. A 3 mm differential laxity threshold between the two knees at a given load represented a complete ACL tear. A 1.5 mm difference was predictive of a partial tear.²³ (Fig. 2).

Fig. 2.

Example of laximetry measurement with GNRB at 134 N: differential >3 mm between the pathological knee (red curve) and the healthy knee (green curve), confirming the diagnosis of complete ACL rupture.

2.5. Statistical analysis

The statistical analysis was performed by a biostatistician using XLSTAT Life Sciences (Addinsoft, Bordeaux, France) software. Variables were analyzed with the Chi-squared test and Monte Carlo simulation. The two-sided Mann-Whitney test with Monte Carlo simulation was used to compare two variables and the Kruskal-Wallis test when there were more than two. Significance level was set at p ≤ 0.05.

2.6. Ethical consideration

The institutional review board granted approval for this retrospective study (No. 2020-024).

3. Results

3.1. Objective evaluation

At a mean 6.3-year follow-up (2–13), the complete retear rate based on the instrumented laxity assessment at 134 N was 6.6 % (10/151). The partial retear rate was 23.2 % (35/151). At 200 N, retear rates increased to 11.9 % (18/151) for complete tears and 22.5 % (34/151) for partial tears. If we add the two patients who underwent revision surgery before the study, the complete retear rate reaches 7.8 % (12/153) at 134 N and 13.1 % (20/153) at 200 N (Fig. 3).

Fig. 3.

Distribution of knee counts based on the measured laximetry differential in millimeters at 134 N and 200 N.

The only significant risk factor at 134 N was sex, with a higher retear rate (complete and partial) in men (p = 0.04). The other preoperative (age, BMI, Tegner) and intraoperative (surgeon experience level, type of graft, associated meniscus injury) risk factors were comparable. Ligament testing in the frontal plane and the quadriceps muscle status at the time of review were the same, whether patients suffered retear or not.

At 200 N, the only risk factors for recurrence were a significantly lower BMI (p = 0.05) and using a HS graft (p = 0.04) (Table 2).

Table 2.

Impact of pre- intra- and postoperative data on the retear rate measured with GNRB® at 134 N and 200 N.

	134 N				200 N
	Normal	Partial	Total	P	Normal	Partial	Total	P
Age (years)	29,4(15; 59)	28,4(15; 53)	26,8(19; 37)	0,73	29,9(15; 59)	26,4(17; 44)	29,2(17; 53)	0,14
Gender				0,04				0,18
Male	78	21	10		74	19	16
Female	28	14	0		25	15	2
Mean BMI (range)	26,1(19; 49)	24,8(18; 40)	25 (20; 30)	0,39	26,1(19; 49)	25,9(19; 40)	23,1(18; 30)	0,05
Tegner score				0,88				0,71
0-1-2	10	6	2		10	5	3
3-4-5-6	81	24	7		75	25	12
7-8-9-10	15	5	1		14	4	3
Surgeon				0,37				0,2
Junior	71	19	6		62	14	12
Senior	35	16	4		32	15	5
Graft				0,25				0,04
Hamstring	32	13	1		30	14	2
BPTB	74	22	9		74	14	17
Meniscal injury				0,31				0,9
No	68	20	4		67	13	12
Yes	39	14	6		38	11	10
Thigh circumference (cm)	−0,7 (−4; 4)	−0,9 (−6; 3)	−0,1 (−3; 1)	0,17	−0,8 (−4; 4)	−0,6 (−6; 3)	−0,6 (3; 1)	0,64

3.2. Subjective functional outcomes (Table 3)

Table 3.

Correlation between patient's subjective instability and functional scores (Lysholm, KOOS, ACL-RSI).

	Stable	Unstable	p
LYSHOLM			<0.0001
excellent	78	6
average	25	13
poor	12	12
KOOS
Total	84 (29.2–100)	71.3 (30.4–95.1)	<0.0001
Symptoms	84 (28.6–100)	75.6 (32.1–100)	0.005
pain	88.9 (38.9–100)	80.4 (33.3–100)	0.001
Daily life activities	92.3 (33.8–100)	85.9 (45.6–100)	<0.0001
Sport	78.5 (10–100)	59.2 (5–100)	<0.0001
Quality of life	76 (0–100)	55.7 (6.3–100)	<0.0001
ACL-RSI	66.4 (5–100)	48.2 (4.2–95)	<0.0001

During the patient interviews, 22.5 % (34/151) of patients felt their operated knees were unstable. This perceived instability was reflected in significantly lower functional scores, including KOOS (total and itemized), Lysholm, and ACL-RSI (0.0001 < p < 0.02), at the review.

3.3. Correlation between objective knee laxity assessment and subjective functional assessment

When patients described their knee as unstable, the mean laxity measured at 134 N and 200 N was significantly worse (p < 0.0001) and was associated with a higher complete retear rate (p = 0.015). However, the total KOOS, Lysholm and ACL-RSI functional scores weren't correlated with the objective knee laxity measured. As well, the time to return to sports, sports level, and patient satisfaction were not correlated with knee laxity. The Lachman test was normal in knees that had a partial tear based on the GNRB.

4. Discussion

The complete retear rate was between 7.8 % (134 N) and 13.1 % (200 N) according to the GNRB (134 N or 200 N). Our failure rate is slightly higher than the average 5.8 % reported in the literature at 134 N, regardless of technique, although some studies have reached 10.4 % with the anteromedial portal technique.⁴ However, the failure rate at 200 N is more than twice as much when we compare our results to those of other published studies. These findings raise concerns, but they are derived from an objective and methodologically rigorous measurement, which surpasses the methodologies used in studies that reported lower retear rates. Most studies assess outcomes using only a subjective functional score or in combination with clinical tests relying on the Rolimeter (Aircast Europa, Neubeuern, Germany) or KT-1000 (MEDmetricCorp, San Diego, USA). Results with these devices are examiner-dependent since force is not quantified but applied manually, whereas the automated GNRB provides reproducible application for all patients. The 3 mm threshold to diagnose a complete retear at 134 N came from previously published studies.³⁰ Considering these observations, a 7.8 % failure rate in this study seems justifiable.

The 23 % partial retear rate at 134 N and 200 N was also important, although difficult to compare to the data in the literature. Indeed, the chosen 1.5 mm GNRB differential laxity threshold cannot be compared with the other laxity tests because they do not have a definitive threshold value for partial tears. Furthermore, clinical or paraclinical partial tear diagnostics are difficult to ascertain. In point of fact, the Lachman test was normal for patients with a partial retear in our study. It is possible the 1.5 mm difference we observed was actually due to the ligaments, and may be linked to graft type, as mean knee laxity outcomes were slightly worse with the HS graft. There is a consideration that these observed partial ruptures may actually be attributed to a slight "release" of the graft rather than true partial ruptures. It remains unclear how to differentiate between these two possibilities. Additionally, the threshold value of 1.5 mm used for assessment raises questions. It is challenging to determine this value manually during clinical examination. The measurement of this threshold value above 1.5 mm is conducted using the GnRB method, which has been validated for diagnosing partial tears (3 mm for complete rupture) of the ACL in preoperative examinations and for determining the indication for ligamentoplasty after trauma. Furthermore, it is also impossible to explain these outcomes through postoperative quadriceps atrophy, or meniscus injury, as they have a known stabilizing effect on the knee. This underscores how several factors are involved in a partial tear diagnosis.³¹

This study reports the outcomes of a technique performed in a University Hospital by experienced physicians and senior residents. Considering that the surgeries are performed by a single, experienced surgeon in most studies, this could be part of the reason for the slightly higher retear rate. Arthroscopic ligament reconstruction is a procedure with a relatively long learning curve and inaccurate tunnel placement. The main cause for ligament reconstruction failure³² is common during the early stages of experience. In fact, this is one of the main drawbacks of the TT technique, where femoral positioning is constrained and therefore less anatomic than with the outside-in technique.^12,33 Optimal femoral tunnel placement requires a minimum of expertise and graft inclination in the sagittal plane may increase the risk of retear.³⁴ Graft ascension through the tunnels may also lead to weakening.³⁵ A radiological study is currently under way to help us understand the causes underlying our failure rate.

Another explanation for the high retear rate may be the choice of primary surgical procedure. Among the studied cohort, no ALL reconstructions had been performed, although there were 12 cases of preoperative rotatory instability that cannot be managed with an isolated ACL reconstruction. Although the tunnels are isometric in the TT technique,15, 16, 17 rotatory instability requires ALL reconstruction using an open or an arthroscopic approach if the outside-in technique is preferred. The outside-in technique, even as a single bundle reconstruction, always provides better rotational knee stability than the TT technique according to Seo et al..¹⁶ Considering that the latest GNRB model, the Dynelax, measures tibial rotation as well as translation, it may be able to identify ALL injury, and help us explain our study's high retear rate.

The subjective functional outcomes also raise some interesting, even surprising questions. The 25 % instability rate reported by patients at the last follow-up was high, reflecting the data from the various functional scores, which weren't not correlated with the knee laxity data. That may mean that when a patient reports knee instability, the surgeon should consider the possibility of a partial or complete ACL retear. This subjective complaint must prevail over poor functional scores, satisfaction levels or return to sports when determining outcomes. Furthermore, they should prompt further investigation (laxity test or MRI) to confirm or rule out a retear.

The limitations of our study pertain to its retrospective and non-comparative design. The retear results may have been worse if the patient follow-up rate had been higher. Among the patients refusing to participate, several did not report reoperations or postoperative problems during the telephone interview. Furthermore, there were no X-ray or CT evaluations conducted to assess the positioning of the tunnels. This lack of imaging data limits our ability to precisely evaluate the anatomical placement of the grafts. Nevertheless, this limitation was counteracted by a rigorous evaluation approach of laxity using the GnRB method, which may differ from other studies that relied on less stringent techniques such as the Rolimeter. Therefore, the study retains its strengths in terms of its objective and robust methodology with GNRB measured outcomes, a sizable patient cohort, and a relatively long mean follow-up period.

5. Conclusion

ACL reconstruction by TT drilling technique will likely remain the reference technique in our department as it offers several validated advantages. The rate of complete retear is comparable to that reported in the literature with other techniques. However, these results should prompt us to improve our indications and expand our range of possible techniques.

CRediT authorship contribution statement

Pauline Unal: Study design, manuscript writing and editing, data acquisition. Ramy Samargandi: data acquisition, manuscript writing and editing. Maxime Saad: manuscript review and editing. Céline Turbillon: Formal analysis, Writing – review & editing. Antoine Schmitt: data acquisition, Writing – review & editing. Julien Berhouet: study design, manuscript writing and editing, Supervision, All authors have read and agreed to the published version of the manuscript.

Informed consent statement

The authors certify that they have obtained all appropriate patients consent.

Institutional review board statement

Ethical approval for this study was obtained from our institutional review board (n° 2020-024).

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of competing interest

Authors have no conflict of interest to declare.