Anatomic Double-Bundle Anterior Cruciate Ligament Reconstruction Using Bone–Patellar Tendon–Bone and Gracilis Tendon Graft: A Comparative Study With 2-Year Follow-up Results of Semitendinosus Tendon Grafts Alone or Semitendinosus–Gracilis Tendon Grafts

Anatomic Double-Bundle Anterior Cruciate Ligament Reconstruction Using Bone–Patellar Tendon–Bone and Gracilis Tendon Graft: A Comparative Study With 2-Year Follow-up Results of Semitendinosus Tendon Grafts Alone or Semitendinosus–Gracilis Tendon Grafts Yasuo Niki, M.D., Ph.D., Hideo Matsumoto, M.D., Ph.D., Akihiro Hakozaki, M.D., Hiroya Kanagawa, M.D., Yoshiaki Toyama, M.D., Ph.D., and Yasunori Suda, M.D., Ph.D.

Purpose: The purpose of this study was to assess the clinical results of anatomic double-bundle anterior cruciate ligament (ACL) reconstruction by use of bone–patellar tendon– bone and gracilis tendon (BPTB-G) grafts and to compare them with the results of double-bundle ACL reconstruction by use of semitendinosus tendon (ST) or semitendinosus-gracilis tendon (ST-G) grafts, with particular emphasis on the postoperative incidence of anterior knee pain. Methods: The study comprised 144 patients who underwent unilateral anatomic double-bundle ACL reconstruction with 3 graft types, including 55 BPTB-G, 56 ST, and 33 ST-G grafts. A traumatic graft rupture occurred within 2 years postoperatively in 5 patients (1 BPTB-G, 3 ST, and 1 ST-G). Clinical results and incidence and severity of anterior knee pain were assessed and compared among the 3 different graft groups at 2 years postoperatively. Potential variables influencing postoperative anterior knee pain development were subjected to univariate analysis, followed by logistic regression analysis to identify risk factors for anterior knee pain. Results: Both subjective and objective clinical results in anatomic double-bundle ACL reconstruction with BPTB-G graft were similar to those using ST or ST-G graft at 2 years postoperatively. The incidences of anterior knee pain at 2 years’ follow-up were 18.5%, 9.4%, and 9.3% in the BPTB-G, ST, and ST-G groups, respectively, indicating no statistically significant difference among the 3 groups. Multivariate logistic regression analyses showed that BPTB graft harvest and patellofemoral cartilage defect failed to be significant factors for anterior knee pain whereas quadriceps peak torque at 60°/s was the only significant factor for anterior knee pain at 2 years. Conclusions: Clinical results including the incidence of anterior knee pain 2 years after anatomic double-bundle ACL reconstruction with BPTB-G grafts were comparable to those after ACL reconstruction with ST or ST-G grafts. Level of Evidence: Level III, therapeutic, retrospective comparative study.

A

From the Department of Orthopaedic Surgery (Y.N., A.H., H.K., Y.T., Y.S.) and Institute for Integrated Sports Medicine (H.M.), Keio University, Tokyo, Japan. The authors report no conflict of interest. Received August 4, 2010; accepted March 25, 2011. Address correspondence to Yasuo Niki, M.D., Ph.D., Department of Orthopaedic Surgery 35, Shinanomachi, Shinjuku-ku, Tokyo, JP, 160-8582 Japan. E-mail: [email protected] © 2011 by the Arthroscopy Association of North America 0749-8063/10465/$36.00 doi:10.1016/j.arthro.2011.03.086

1242

nterior cruciate ligament (ACL) reconstruction is currently one of the most common surgical procedures in sports medicine, and it is generally accomplished arthroscopically by use of bone–patellar tendon– bone (BPTB) or hamstring tendon (HT) autografts. During the last decade, there has been an increased use of HT graft, because of the lower incidence of postoperative complications, particularly fewer cases of donor-site morbidity. Indeed, recent systematic reviews of published randomized clinical trials comparing BPTB and HT autografts have re-

Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 27, No 9 (September), 2011: pp 1242-1251

GRAFT CHOICE IN DOUBLE-BUNDLE ACL RECONSTRUCTION ported that BPTB graft yielded better stability; however, BPTB graft was inferior to HT graft in terms of postoperative complications, which is 1 of the reasons that BPTB graft has decreased in popularity.1,2 There are a substantial number of postoperative complications after primary ACL reconstruction, including range-of-motion deficit, quadriceps weakness, and donor-site morbidity, particularly associated with ACL reconstruction with the BPTB graft. Donor-site morbidity manifests clinically as anterior knee pain, donor-site tenderness, pain on kneeling, and loss of anterior knee sensitivity. Of these, anterior knee pain is a frequent and important complication with a potential to impede rehabilitation and return to sports activity, and anterior knee pain lowers functional knee scores such as the Lysholm score3 and International Knee Documentation Committee (IKDC) rating scale.4 A previous study has advocated that the size of residual bony defect after harvesting BPTB graft was closely related to the development of anterior knee pain after ACL reconstruction with the BPTB graft.5 We have developed an alternative ACL reconstruction technique that involves an anatomic double-bundle method using a unique combination of BPTB and gracilis tendon autografts. This technique uses a relatively small BPTB graft (8 mm in diameter) for the anteromedial bundle (AMB) and gracilis tendon for the posterolateral bundle (PLB). We believe that this graft combination can minimize donor-site morbidity, by reducing the diameter of BPTB graft and leaving the sartorius tendon and semitendinosus tendon (ST) intact. To date, multiple factors besides the graft choice have been reported to increase the risk for postoperative anterior knee pain, including postoperative extension deficit,3,6-12 flexion deficit,3,6,11,12 residual anterior instability,5 gender,6 age,11 and body mass index. Given the multifactorial nature of anterior knee pain development after ACL reconstruction, multivariate analysis should be used to investigate the factors affecting anterior knee pain. The first purpose of this study was to assess the clinical results of our anatomic double-bundle ACL reconstruction using a unique graft combination, BPTB and gracilis tendon (BPTB-G), and to compare them with the results of double-bundle ACL reconstruction with the HT graft (either ST alone or semitendinosus-gracilis tendon [ST-G] grafts), with particular emphasis on postoperative anterior knee pain. The second purpose was to elucidate the risk factors for anterior knee pain 2 years after anatomic double-bundle reconstruction by use of logistic regression analy-

1243

sis. Our hypotheses were that the clinical results, including the incidence of anterior knee pain, after ACL reconstruction with BPTB-G graft would be similar to those after ACL reconstruction with ST or ST-G graft and that a harvest of relatively small BPTB graft would not be a risk factor for anterior knee pain development at 2 years postoperatively.

METHODS Patients From August 2006 to June 2008, a total of 156 consecutive patients (159 knees) underwent primary anatomic double-bundle ACL reconstruction at our institute. Inclusion criteria were follow-up of 2 years or more and full clinical data available at 2 years postoperatively, including clinical scores, anteroposterior stability, and thigh muscle strength. Exclusion criteria comprised previous ligament reconstruction, multiple ligament injuries, bilateral ACL injuries, and concomitant treatments for articular cartilage defects, such as osteochondral autologous transplantation and microfracture. Of the 156 patients, 3 had bilateral ACL insufficiency, 4 were lost to follow-up, 2 had grade III medial collateral ligament injury, and 3 simultaneously underwent treatment for articular cartilage defects (1 osteochondral autologous transplantation and 2 microfractures). Therefore 144 patients met our inclusion criteria and were enrolled in this study. Patient characteristics are shown in detail in Table 1. Regarding graft selection, our cohort received 55 BPTB-G grafts, 56 ST grafts, and 33 ST-G grafts. Use of the 3 grafts was allocated based on when the operation was performed, with BPTB-G mainly used between August 2006 and June 2007 and ST or ST-G mainly used between July 2007 and June 2008. In some small female patients, reconstruction with BPTB-G was considered contraindicated, so they underwent reconstruction with ST or ST-G, because the BPTB graft for the AMB potentially disturbs bone tunnel creation for the PLB in small knees. Associated meniscal treatments and articular cartilage defects based on Outerbridge classification are summarized in Table 1. Overall, the baseline characteristics of the patients, except for gender and follow-up period, did not differ significantly among the 3 groups. In terms of gender difference in graft selection, ST-G graft tended to be used in female patients because ST tendon in female patients has shorter length and small diameter than that in male patients.

1244

Y. NIKI ET AL. TABLE 1.

Demographic Characteristics of Patients

Variable No. of cases/No. of knees Age (range) (y) Gender (M/F) Follow-up period (range) (y) Time from injury to index operation (range) (y) Preinjury Tegner activity level (range) Preoperative Lysholm score Treatment for meniscal lesion (No. of cases of partial resection/repair) MM LM Articular cartilage defect (Outerbridge classification II/III/IV) PF MFC MTP LFC LTP

BPTB-G Group

ST Group

ST-G Group

55/55 28.4 (14-46) 37/18 3.4 (2.0-4.2) 3.4 (0.1-28) 6.0 (3-9) 72.6 (52-90)

56/56 28.7 (14-49) 22/34 2.5 (2.0-3.2) 3.0 (0.1-20) 6.2 (3-9) 68.8 (43-90)

33/33 30.5 (14-52) 5/28 2.4 (2.0-3.0) 3.2 (0.1-25) 5.2 (2-8) 67.0 (35-90)

9/2 7/2

1/1/0 4/1/1 2/0/0 2/1/1 2/0/0

7/2 11/1

2/1/0 3/3/1 2/1/0 1/1/0 1/0/0

Significance NS P ⬍ .0001* P ⬍ .0001* NS NS NS NS

10/2 6/0

1/1/0 1/2/1 1/0/1 0/0/0 0/0/0

NS NS NS NS NS

Abbreviations: NS, not significant; MM, medial meniscus; LM, lateral meniscus; PF, patellofemoral; MFC, medial femoral condyle; MTP, medial tibial plateau; LFC, lateral femoral condyle; LTP, lateral tibial plateau. *Statistically significant.

Surgical Techniques All ACL reconstructions were arthroscopically performed by a single surgeon using anatomic doublebundle procedures. All surgeries were performed with patients under general anesthesia and with tourniquets. As reported previously,13 BPTB was used as the AMB graft and gracilis tendon was used as the PLB graft in BPTB-G double-bundle reconstruction. During BPTB graft preparation, the central third of the ipsilateral patellar tendon was harvested as a uniform graft 8 mm in diameter, with bone plugs of 2 cm in length and 8 to 9 mm in diameter on each side. Use of a bone plug with a diameter greater than 9 mm should be avoided to avoid disturbing the anatomic footprint of the PLB and to maintain the site for creation of a bone tunnel for the PLB. For the PLB graft, the gracilis tendon was harvested and folded as a doubled graft with EndoButton CL (Smith & Nephew Endoscopy, Andover, MA) of appropriate length placed at the loop end. A baseball-glove suture using No. 2-0 FiberWire (Arthrex, Naples, FL) was used on both free ends of the graft. In HT double-bundle ACL reconstruction, 2 double-looped ST tendons were prepared for both the AMB and PLB grafts when the ST tendon was 24-cm long or more. Gracilis tendon was harvested for the PLB graft and double looped only when the ST tendon was less than 24 cm long. After

we removed the ACL remnant and exposed the bony ridge of the medial wall of the lateral femoral condyle (i.e., resident’s ridge), two 2.4-mm guidewires were inserted into both the AMB and PLB attachments in an inside-out fashion through a medial accessory portal (i.e., far anteromedial [AM] portal) that was routinely created. Our landmarks for ideal AMB and PLB bone tunnels were the resident’s ridge, which reportedly represents the anterior border of the femoral ACL footprint and is identifiable in more than 97% of human femora.14,15 In the sagittal plane, tunnels for the AMB and PLB were placed posterior to the resident’s ridge, and the centers of these tunnels corresponded to the center of each anatomic footprint, including the fanlike portion.14,16 With such placement, the bone tunnel for the AMB was normally placed 3 mm or less from the posterior edge of the notch (Fig 1). Bone tunnels for the AMB and PLB were drilled over the guidewire when the knee was flexed greater than 120°. The diameter of the bone tunnel for the PLB was normally 5 to 6 mm, regardless of graft type, whereas the bone tunnel for the AMB was 8 or 9 mm in diameter for BPTB graft and 5 to 7 mm for HT graft. During tibial bone tunnel creation, two 2.4-mm guidewires were inserted into the center of the AMB and PLB footprints by use of a drill guide system (Smith & Nephew Endoscopy).

GRAFT CHOICE IN DOUBLE-BUNDLE ACL RECONSTRUCTION

1245

FIGURE 1. (A) Arthroscopic image with knee at 90° of flexion showing femoral tunnel locations for AMB and PLB in anatomic doublebundle ACL reconstruction with BPTB-G. (B) BPTB and gracilis tendon grafts were secured in the bone tunnels for the AMB and PLB, respectively.

Each bone tunnel was overdrilled, similar to the femoral bone tunnel creation. The grafts were introduced through the tibial tunnel to the femoral tunnel, and femoral fixations for ST or G tendon grafts were achieved by use of EndoButton CL (Smith & Nephew Endoscopy) whereas BPTB graft for the AMB was fixed with EndoButton CL BPTB (Smith & Nephew Endoscopy) in BPTB-G reconstructions. Tibial graft fixations were achieved with a double-spike plate (DSP; Meira, Nagoya, Japan) for ST or G tendon grafts and with an interference screw (Soft Silk; Smith & Nephew Endoscopy) for BPTB graft. After in situ pre-tensioning was performed by loading 30 N each on the AMB and PLB grafts, both grafts were secured with 20 N of pretension with the knee at 20° of flexion by use of a ligament tensioner (Smith & Nephew Endoscopy). Postoperative Rehabilitation The knee was immobilized for 3 days, followed by quadriceps isometric and passive flexion exercises. Patients were allowed one-third body weight bearing 2 weeks postoperatively, increasing to full weight bearing as tolerated by pain and effusion at 4 weeks postoperatively. Running was allowed at 3 months postoperatively, followed by return to cutting actions and contact sports at 9 months postoperatively or later. Evaluation Follow-up examinations were performed at 2 years postoperatively, and the data were analyzed according to the type of graft used (BPTB-G, ST, or ST-G). The data were compared among the 3 graft types in terms of Tegner activity level, anteroposterior knee laxity measured with a KT-2000 arthrometer (MEDmetric, San Diego, CA), pivot-shift test (graded as equal, glide, clunk, and gross), Lysholm score, IKDC eval-

uation, knee extension and flexion deficit measured with a goniometer, and quadriceps strength. These parameters were compared between the BPTB-G, ST, and ST-G groups. In the IKDC rating system, each parameter was graded as normal, nearly normal, abnormal, or severely abnormal, with the lowest gradation within a group determining the group gradation and the worst gradation within a group determining the final evaluation. Isokinetic peak extension torque at 60°/s was measured with an isokinetic dynamometer (Biodex Medical Systems, Shirley, NY), for use as an indicator of quadriceps strength. Assessment of Anterior Knee Pain Anterior knee pain was diagnosed when a patient undergoing BPTB-G reconstruction presented with soreness at the harvest site or when a patient undergoing ST or ST-G reconstruction presented with soreness and tenderness at the inferior pole of the patella. Anterior knee pain normally occurred during either double- or single-leg squat training and was reproducible on single-leg squat during clinical assessment in the outpatient clinic. The incidence and severity of anterior knee pain were evaluated at 2 years postoperatively and compared between the BPTB-G, ST, and ST-G groups. Severity of anterior knee pain was classified into 3 stages as follows: I, pain after activity only; II, pain during and after activity but still able to perform at a satisfactory level; and III, pain during and after activity and more prolonged and unable to perform at a satisfactory level. After patients were divided into 2 groups based on the presence or absence of anterior knee pain, univariate analysis and subsequent multivariate logistic regression analysis were performed to determine factors significantly affecting anterior knee pain.

1246

Y. NIKI ET AL. TABLE 2.

Clinical Results of 3 Different Double-Bundle ACL Reconstructions at 2 Years Postoperatively

ACL reinjury (No. of cases) Lysholm score Tegner activity level IKDC (No. of cases) A (normal) B (nearly normal) C (abnormal) D (severely abnormal) Pivot-shift test (No. of cases) Equal Glide Clunk Gross Anteroposterior laxity* Mean (range) (mm) ⱕ2 mm (%) Knee extension deficit (No. of cases)† None ⱕ5° 6°-10° Knee flexion deficit (No. of cases)† None ⱕ5° 6°-15° Quadriceps torque at 60°/s‡

BPTB-G Group (n ⫽ 55)

ST Group (n ⫽ 56)

ST-G Group (n ⫽ 33)

Significance

1

3

1

NS

94.7 ⫾ 5.5 5.4 ⫾ 1.3

96.2 ⫾ 5.1 5.9 ⫾ 1.2

95.8 ⫾ 5.9 5.8 ⫾ 1.7

NS NS NS

28 18 6 2

27 21 4 1

18 10 3 1 NS

49 3 1 1

46 3 3 1

27 2 2 1

0.8 ⫾ 1.7 (–3.5 to 4) 82

1.0 ⫾ 1.7 (–2 to 5) 78

1.1 ⫾ 1.4 (–1 to 5) 84

51 1 2

47 4 2

29 2 1

NS NS NS

NS 48 4 2 85.6 ⫾ 12

46 5 2 88.5 ⫾ 10

29 1 2 87.6 ⫾ 12

NS

Abbreviation: NS, not significant. *Side-to-side difference at 134 N torque, as measured by KT-2000. †Extension and flexion deficit expressed as difference compared with contralateral side. ‡Values are expressed as percent of healthy side.

Statistical Analysis Statistical analysis was performed with SPSS software, version 17.0 (SPSS, Chicago, IL). Comparison between the 3 different graft groups was conducted by use of 1-way analysis of variance with post hoc testing by the Bonferroni method. The ␹2 test and Fisher exact test were used to compare categorical variables between the groups. Statistical significance was defined as P ⬍ .05. For univariate analysis of continuous variables between patients with and without anterior knee pain, an unpaired Student t test was performed. After determining variables associated with anterior knee pain at the

P ⬍ .4 level according to univariate analysis, multiple logistic regression analysis was performed. Statistical significance for the multivariate model was accepted at the P ⬍ .05 level. Sample size estimation indicated that a total of 559 patients would be required to show a statistical difference in anterior knee pain incidence among the 3 groups with an ␣ level of 0.05 and a ␤ level of 80%. When a comparison of quadriceps peak torque was made between the group with anterior knee pain and the group without anterior knee pain, 15 patients in each group would be required to show a difference with an ␣ level of 0.05 and a ␤ level of 80%.

GRAFT CHOICE IN DOUBLE-BUNDLE ACL RECONSTRUCTION TABLE 3. Comparison of Incidence and Severity of Anterior Knee Pain Among 3 Different Double-Bundle ACL Reconstructions BPTB-G ST ST-G Significance Group Group Group (P Value) Anterior knee pain incidence 10 (18.5) 5 (9.4) 3 (9.3) NS (.29) [No. of cases (%)] Anterior knee pain severity (No. of cases) Stage I 6 4 3 Stage II 3 1 0 Stage III 1 0 0 Abbreviation: NS, not significant.

RESULTS During the 2-year follow-up period, 1 patient in the BPTB-G group, 3 in the ST group, and 1 in the ST-G group had a traumatic graft rupture and are reported as ACL reinjury cases (Table 2). Single-bundle revision ACL reconstruction was performed with contralateral BPTB graft in 1 of these patients and ipsilateral BPTB graft in 4. These 5 patients were excluded from the calculations. There was no major postoperative complication in any group, except for 1 deep venous thrombosis, 1 pulmonary embolism, and 1 case of pseudogout. Two patients in the BPTB-G group and two in the ST group underwent additional meniscus surgery during the follow-up periods. These 4 patients displayed 4 mm of anteroposterior instability or greater as measured by KT2000 arthrometer, with a positive pivot-shift test graded as glide or higher. At 2 years follow-up, there were no significant differences in terms of the Lysholm score, Tegner activity level, and IKDC evaluation among the 3 groups. There were 2 patients evaluated as having IKDC rank D in the BPTB-G group, comprising 1 patient with grade III anterior knee pain and 1 with grade II anterior knee pain. There were 2 patients with rank C in the BPTB-G group, both of whom had grade II anterior knee pain. The Lysholm score and IKDC evaluation—representing subjective and objective clinical measures, respectively— improved significantly from preoperatively to 2 years follow-up in all 3 groups. On KT-2000 measurement of anteroposterior knee laxity, the mean side-to-side difference under 134 N was 0.8 ⫾ 1.7 mm, 1.0 ⫾ 1.7 mm, and 1.1 ⫾ 1.4 mm in the BPTB-G, ST, and ST-G groups, respectively. The BPTB-G group represented slightly good stability, but no statistically significant differences were found among the 3 groups. The percentage of patients

1247

with 2 mm of anteroposterior laxity or less was 82%, 78%, and 84% in the BPTB-G, ST, and ST-G groups, respectively. These data represented satisfactory results and did not show statistically significant differences among the 3 groups. Regarding the rotatory instability measured by the manual pivot-shift test, 5 patients in the BPTB-G group, 7 in the ST group, and 5 in the ST-G group had positive findings, which did not yield a statistically significant difference. Concerning the postoperative loss of terminal knee motion and mean quadriceps peak torque at 60°/s, there was no significant difference among the 3 groups. Incidences of anterior knee pain at 2 years followup were 18.5%, 9.4%, and 9.3% in the BPTB-G, ST, and ST-G groups, respectively (Table 3). Most patients presenting with anterior knee pain complained of harvestsite pain during either double- or single-leg squat training, with anterior knee pain reproduced on single-leg squat in the outpatient clinic. Actually, the incidences of overall anterior knee pain markedly decreased with time and averaged 12.9% at 2 years follow-up; however, the BPTB-G group still showed a high incidence of anterior knee pain as compared with the ST group or ST-G group, although there was no statistically significant difference between the groups (P ⫽ .29). Factors potentially affecting anterior knee pain development during 2 years’ follow-up were compared between the patients with anterior knee pain and those without anterior knee pain (Table 4). According to the univariate analysis, both BPTB graft harvest and the presence of patellofemoral cartilage defect, reportedly considered to be causative factors for anterior knee pain, were not found to be associated with anterior knee pain development. Quadriceps peak torque at 60°/s was the only factor significantly associated with anterior knee pain. Multivariate logistic regression analyses—with preinjury Tegner activity, time from injury to index operation, harvest of BPTB graft, patellofemoral cartilage defect, and quadriceps peak torque at 60°/s included as covariates—were performed to predict anterior knee pain development at 2 years postoperatively (Table 5). Quadriceps peak torque at 60°/s was the only factor found to be inversely associated with risk for anterior knee pain development at 2 years postoperatively (odds ratio, 0.94; 95% confidence interval, 0.89 to 0.98). BPTB graft harvest and patellofemoral cartilage defect were not significant factors for anterior knee pain development, which corresponded to the results of univariate analysis.

1248

Y. NIKI ET AL. TABLE 4.

Univariate Analysis of Risk Factors for Anterior Knee Pain Development at 2 Years Postoperatively Group With Anterior Knee Pain (n ⫽ 18)

Group Without Anterior Knee Pain (n ⫽ 121)

28.7 9/9 61.2 5.5

29.1 62/59 39.5 5.9

NS(.87) NS(.92) NS(.33) NS(.35)

1.0 (1.9) 77.8 73.3 10 (56) 2

0.9 (1.6) 83.5 71.2 44 (36) 5

NS(.81) NS(.88) NS(.73) NS(.12) NS(.19)

2 3 77.1 ⫾ 12

10 13 88.7 ⫾ 10

Age (y) Gender (M/F) Time from injury to index operation (mo) Preinjury Tegner activity level Anteroposterior laxity* (mm) Mean (SD) ⱕ2 mm (%) Preoperative Lysholm score (mean) Harvest of BPTB graft [No. of cases (%)] PF cartilage defect (No. of cases) (Outerbridge grade II or greater) Knee extension deficit (No. of cases) Knee flexion deficit (No. of cases) Quadriceps torque at 60°/s†

Significance (P Value)

NS(.68) NS(.46) P ⫽ .03‡

Abbreviations: NS, not significant; PF, patellofemoral. *Side-to-side difference at 134 N torque, as measured by KT-2000. †Values are expressed as percent of healthy side. ‡Statistically significant.

DISCUSSION Anterior knee pain has been reported to impair functional recovery after ACL reconstruction and is considered 1 of the most essential targets of evaluation when assessing the clinical results of ACL reconstruction. Indeed, certain authors have reported that clinical results in patients with anterior knee pain were inferior to those in patients without anterior knee pain when TABLE 5. Multivariate Logistic Regression Analysis of Risk Factors for Anterior Knee Pain at 2 Years Postoperatively Risk Factors for Anterior Knee Pain

Odds Ratio

95% Confidence Interval

Preinjury Tegner activity Time from injury to index operation Harvest of BPTB graft PF cartilage defect (Outerbridge grade II or greater) Quadriceps torque at 60°/s*

0.77

0.51-1.16

.22

1.00

0.99-1.01

.29

3.24

0.95-11.0

.06

1.36

0.10-18.5

.82

0.94

0.89-0.98

.007†

Abbreviation: PF, patellofemoral. *Values are expressed as percent of healthy side. †Statistically significant.

Significance (P value)

measured using the Lysholm score3 and IKDC rating scale.4 Various factors potentially contribute to postoperative anterior knee pain development at long-term follow-up, but most studies to date have focused on anterior knee pain associated with donor-site morbidity, particularly after ACL reconstruction with BPTB graft. In fact, according to our comparative analysis of the BPTB-G group versus the ST or ST-G group, the incidence of anterior knee pain in the BPTB-G group was 58.2% (32 of 55 cases) at 2 months follow-up and was significantly higher than that in the ST or ST-G group; however, the rate of anterior knee pain fell to 18.5% (10 of 54 cases) at 2 years follow-up (Table 3). Such an early impact of donor-site morbidity on anterior knee pain was consistent with the findings reported by Feller et al.17 Multivariate logistic regression analysis showed that low quadriceps muscle power but not donor-site morbidity was associated with anterior knee pain development at 2 years postoperatively, although it still remains unclear whether reduced quadriceps torque was a cause or consequence of anterior knee pain. It is presumed that the harvest of relatively small BPTB graft in our procedure using the BPTB-G graft combination contributed to the reduced influence of donor-site morbidity on anterior knee pain at 2 years follow-up, because a previous biomechanical study has shown that bone plug removal from the patella increased axial strain of the patella and abrogated stress distribution on the

GRAFT CHOICE IN DOUBLE-BUNDLE ACL RECONSTRUCTION patellar surface and that these adverse effects were restored by filling of the bony defect.18 Moreover, from a clinical standpoint, Tsuda et al.5 have reported that residual patellar bony defect (⬎2 mm) after the BPTB graft harvest was significantly associated with anterior knee pain. To better replicate normal knee kinematics and control rotational instability, ACL reconstruction has increasingly focused on the anatomic reconstruction of the AMB and PLB by use of a variety of graft choices and fixation options.19-24 Recent clinical studies on anatomic double-bundle reconstruction have shown better restoration of knee stability than with single-bundle reconstruction.25-29 Most of the anatomic double-bundle reconstructions to date have used soft-tissue grafts, including ST or ST-G autografts or occasionally allografts, by virtue of their greater length and strength and easy handling when they are passed into 2 separate tunnels. However, current double-bundle reconstruction techniques using soft-tissue grafts have several potential disadvantages. Creating both the AM and posterolateral (PL) tunnels within the anatomic footprint increases the risk of bony bridge fracture at the apertures of the 2 tunnels, particularly in small knees.30 Critically short tunnels and posterior tunnel wall blowout are reportedly characteristic disadvantages when the femoral tunnel is created with the AM portal technique.31,32 Moreover, in the case of a large knee in which the distance between the AM portal and the center of the footprint is long, the tunnels for the AMB and PLB tend to be parallel but not divergent, and thus tendon-to-tendon healing may easily occur when the bony bridge between the tunnels is broken. This situation may produce more problems at revision ACL reconstruction. Actually, in our case series, 3 revision surgeries after graft rupture were inevitably performed with a single-bundle procedure by use of ipsilateral BPTB graft, because of apparent tunnel communication. Considering these potential risks for a short socket and posterior cortex blowout, as well as the parallel tunnel placement, our unique BPTB-G graft combination may have a great advantage whereby AM tunnels can be secured by bone plugs. Our BPTB-G reconstruction not only provides strong initial stability of the graft but also remains open to future anatomic double-bundle revision of both the AMB and PLB, even though the AM portal technique is used. In fact, our ACL reconstruction using BPTB-G adopted the AM portal technique, which creates the femoral bone tunnels through the far AM portal, with placement at 2 cm posteromedial to the standard AM portal. In this technique the AMB and PLB graft bending angles are smaller than in the transtibial tech-

1249

nique,33 which results in lower stress on the graft at the femoral tunnel aperture and therefore may reduce graft damage. Moreover, this technique allows the femoral tunnel to be placed in a more oblique and more anatomic position, which provides increased rotational stability. Numerous authors have stated that loss of extension causes anterior knee pain,1,3-9 and Shelbourne and Trumper emphasized the importance of regaining full hyperextension to avoid anterior knee pain development.10 In anatomic double-bundle reconstruction, loss of terminal extension is easily caused by excessive initial tension to and shallower tunnel location of PLB graft in the early postoperative phase. Actually, in our cohort 43% of patients (62 of 144) presented with terminal extension loss and this was significantly relevant to the incidence of anterior knee pain at 2 months follow-up, whereas the incidence dropped to 8.6% (12 of 139) by 2 years postoperatively. Moreover, considering the increased rotational stability of the double-bundle ACL reconstruction proven in an experimental cadaveric study,34 excessive tension to the PLB graft may overconstrain tibial rotation and negatively affect the patellofemoral joint, potentially leading to anterior knee pain in the future. Although our case series included small numbers of chondral lesions (Outerbridge grades II or higher) in the patellofemoral joint preoperatively and the incidence of the lesions failed to show statistically significant differences between the group with anterior knee pain and the group without anterior knee pain at 2 years follow-up, care should be taken to secure the PLB graft in the correctly positioned tunnel with appropriate initial tension during double-bundle ACL reconstruction, particularly in patients with patellofemoral chondral lesions. Some limitations must be taken into consideration for this study. First, the relatively small sample size and short duration of follow-up might obscure precise long-term clinical outcomes and the fate of anterior knee pain. Because approximately 18% of patients showed anterior knee pain at 2 years postoperatively after BPTB-G double-bundle reconstruction, the fate of anterior knee pain in these patients should be examined from a long-term perspective. Second, in terms of graft choice, selection bias may have been present in this study, because patient allocation to BPTB-G, ST, or ST-G graft was not randomized and was instead based on the time when the operation was performed. Further studies with randomization of graft types and longer follow-up are warranted to precisely evaluate the clinical utility of the BPTB-G graft combination in anatomic double-bundle ACL reconstruction. In addition, it will be the subject of a

1250

Y. NIKI ET AL.

future study whether this unique graft combination can overcome the potential shortcomings of anatomic double-bundle ACL reconstruction using soft-tissue graft such as ST and ST-G. CONCLUSIONS Clinical results including the incidence of anterior knee pain 2 years after anatomic double-bundle ACL reconstruction by use of a unique BPTB-G graft combination were comparable to those after ACL reconstruction with ST or ST-G graft. REFERENCES 1. Samuelsson K, Andersson D, Karlsson J. Treatment of anterior cruciate ligament injuries with special reference to graft type and surgical technique: An assessment of randomized controlled trials. Arthroscopy 2009;25:1139-1174. 2. Li S, Su W, Zhao J, et al. A meta-analysis of hamstring autografts versus bone-patellar tendon-bone autografts for reconstruction of the anterior cruciate ligament. Knee. 2010 Sep 15. [Epub ahead of print.] 3. Jarvela T, Kannus P, Jarvinen M. Anterior knee pain 7 years after an anterior cruciate ligament reconstruction with a bonepatellar tendon-bone autograft. Scand J Med Sci Sports 2000; 10:221-227. 4. Otto D, Pinczewski LA, Clingeleffer A, Odell R. Five-year results of single-incision arthroscopic anterior cruciate ligament reconstruction with patellar tendon autograft. Am J Sports Med 1998;26:181-188. 5. Tsuda E, Okumura Y, Ishibashi Y, Otsuka H, Toh S. Techniques for reducing anterior knee symptoms after anterior cruciate ligament reconstruction using a bone-patella tendonbone autograft. Am J Sports Med 2001;29:450-456. 6. Aglietti P, Buzzi R, D’Andria S, Zaccherotti G. Patellofemoral problems after intraarticular anterior cruciate ligament reconstruction. Clin Orthop Relat Res 1993:195-204. 7. Bach BR, Jones GT, Sweet FA, Hager CA. Arthroscopyassisted anterior cruciate ligament reconstruction using patellar tendon substitution. Two- to four-year follow-up results. Am J Sports Med 1994;22:758-767. 8. Harner CD, Irrgang JJ, Paul J, Dearwater S, Fu FH. Loss of motion after anterior cruciate ligament reconstruction. Am J Sports Med 1992;20:499-506. 9. Sachs RA, Daniel DN, Stone ML, Garfein RF. Patellofemoral problems after anterior cruciate ligament reconstruction. Am J Sports Med 1989;17:760-765. 10. Shelbourne KD, Trumper R. Preventing anterior knee pain after anterior cruciate ligament reconstruction. Am J Sports Med 1997;25:41-47. 11. Stapleton TR. Complications in anterior cruciate ligament reconstructions with patellar tendon grafts. Sports Med Arthrosc 1997;5:156-162. 12. Kartus J, Magnusson L, Stener S, et al. Complications following arthroscopic anterior cruciate ligament reconstruction. A 2-5-year follow-up of 604 patients with special emphasis on anterior knee pain. Knee Surg Sports Traumatol Arthrosc 1999;7:2-8. 13. Niki Y, Matsumoto H, Enomoto H, Toyama Y, Suda Y. Single-stage ACL revision with BPTB: A case control series of revision of failed synthetic ACL reconstructions. Arthroscopy 2010;26:1058-1065.

14. Ferretti M, Ekdahl M, Shen W, Fu FH. Osseous landmarks of the femoral attachment of the anterior cruciate ligament: An anatomic study. Arthroscopy 2007;23:1218-1225. 15. Farrrow LD, Chen MR, Cooperman DR, Victoroff BN, Goodfellow DB. Morphology of the femoral intercondylar notch. J Bone Joint Surg Am 2007;89:2150-2155. 16. Hara K, Mochizuki T, Sekiya I, et al. Anatomy of normal human anterior cruciate ligament attachments evaluated by divided small bundles. Am J Sports Med 2009;37:2386-2391. 17. Feller JA, Webster KE, Gavin B. Early post-operative morbidity following anterior cruciate ligament reconstruction: Patellar tendon versus hamstring graft. Knee Surg Sports Traumatol Arthrosc 2001;9:260-266. 18. Sharkey NA, Donahue SW, Smith TS, Bay BK, Marder RA. Patellar strain and patellofemoral contact after bone-patellar tendon-bone harvest for anterior cruciate ligament reconstruction. Arch Phys Med Rehabil 1997;78:256-263. 19. Hara K, Kubo T, Suginoshita T, Shimizu C, Hirasawa Y. Reconstruction of the anterior cruciate ligament using a double bundle. Arthroscopy 2000;16:860-864. 20. Kim SJ, Jung KA, Song DH. Arthroscopic double-bundle anterior cruciate ligament reconstruction using autogenous quadriceps tendon. Arthroscopy 2006;22:797.e1-797.e5. Available online at www.arthroscopyjournal.org. 21. Marcacci M, Molgora AP, Zaffagnini S, Vascellari A, Lacono F, Presti ML. Anatomic double-bundle. Anterior cruciate ligament reconstruction with hamstrings. Arthroscopy 2003;19: 540-546. 22. Muneta T, Sekiya I, Yagishita K, et al. Two-bundle reconstruction of the anterior cruciate ligament using semitendinosus tendon with Endobuttons: Operative technique and preliminary results. Arthroscopy 1999;15:618-624. 23. Shino K, Nakata K, Nakamura N. Anatomic anterior cruciate ligament reconstruction using two double-looped hamstring tendon grafts via twin femoral and triple tibial tunnels. Oper Tech Orthop 2005;15:130-134. 24. Yasuda K, Kondo E, Ichiyama H, et al. Anatomic reconstruction of the anteromedial and posterolateral bundles of the anterior cruciate ligament using hamstring tendon grafts. Arthroscopy 2004;20:1015-1025. 25. Järvelä T, Moisala AS, Sihvonen R, et al. Double-bundle. Anterior cruciate ligament reconstruction using hamstring autografts and bioabsorbable interference screw fixation: Prospective randomized, clinical study with 2-year results. Am J Sports Med 2008;36:290-297. 26. Kondo E, Yasuda K, Azuma H, Tanabe Y, Yagi T. Prospective clinical comparisons of anatomic double-bundle versus singlebundle anterior cruciate ligament reconstruction procedures in 328 consecutive patients. Am J Sports Med 2008;36:16751687. 27. Muneta T, Koga H, Mochizuki T, et al. A prospective randomized study of 4-strand semitendinosus tendon anterior cruciate ligament reconstruction comparing single-bundle and double-bundle techniques. Arthroscopy 2007;23:618-628. 28. Siebold R, Dehler C, Ellert T. Prospective randomized comparison of double-bundle versus single-bundle anterior cruciate ligament reconstruction. Arthroscopy 2008;24:137-145. 29. Yasuda K, Kondo E, Ichiyama H, Tanabe Y, Tohyama H. Clinical evaluation of anatomic double-bundle anterior cruciate ligament reconstruction procedure using hamstring tendon grafts: Comparisons among 3 different procedures. Arthroscopy 2006;22:240-251. 30. Zantop T, Kubo S, Petersen W, Musahl V, Fu FH. Current techniques in anatomic anterior cruciate ligament reconstruction. Arthroscopy 2007;23:938-947. 31. Bedi A, Raphael B, Maderazo A, Pavlov H, Williams RJ III. Transtibial versus anteromedial portal drilling for anterior cruciate ligament reconstruction: A cadaveric study of femoral tunnel length and obliquity. Arthroscopy 2010;26:342-350.

GRAFT CHOICE IN DOUBLE-BUNDLE ACL RECONSTRUCTION 32. Lubowitz JH. Anteromedial portal technique for the anterior cruciate ligament femoral socket: Pitfalls and solutions. Arthroscopy 2009;25:95-101. 33. Nishimoto K, Kuroda R, Mizuno K, et al. Analysis of the graft bending angle at the femoral tunnel aperture in anatomic double bundle anterior cruciate ligament reconstruction: A

1251

comparison of the transtibial and the far anteromedial portal technique. Knee Surg Sports Traumatol Arthrosc 2009;17:270-276. 34. Markolf KL, Park S, Jackson SR, McAllister DR. Anteriorposterior and rotatory stability of single and double-bundle anterior cruciate ligament reconstructions. J Bone Joint Surg Am 2009;91:107-118.