nMOVS

a

pR

2

Femoral Tunnel Length in Primary Anterior Cruciate LigamentReconstruction Using an Accessory Medial PortalMarc Tompkins, M.D., Matthew D. Milewski, M.D., Eric W. Carson, M.D.,

Stephen F. Brockmeier, M.D., Joshua C. Hamann, M.D., Joseph M. Hart, Ph.D.,and Mark D. Miller, M.D.

Purpose: The purpose of this study was to evaluate tunnel length during independent femoral tunnel drilling using anaccessory medial portal with the knee in maximal hyperflexion, and correlate the tunnel length and flexion angle withanthropometric data. Methods: During a 1-year period, 106 consecutive patients undergoing primary anterior cruciateligament (ACL) reconstruction were included in the study. All patients underwent independent femoral tunnel drillingusing an accessory medial portal with maximal knee hyperflexion. Tunnel length and maximal intraoperative kneeflexion angles were measured. Additionally, height, weight, and body mass index (BMI), plus the width and depth of thelateral femoral condyle (LFC), were recorded to correlate with tunnel length and knee flexion angles. Results: Averagetunnel length was 37.0 � 3.3 mm (range, 26 to 45), with all but one tunnel greater than 30 mm. Average knee flexionangle was 134.4 � 5.0� (range, 122� to 147�). Height (r ¼ 0.5, P < .001) and weight (r ¼ 0.33, P ¼ .001), but not BMI (r ¼0.14, P ¼ .17), correlated positively with tunnel length. Width (r ¼ 0.46, P < .001) and depth (r ¼ 0.38, P < .001) of theLFC also correlated positively with tunnel length. Knee flexion angle was not correlated with tunnel length (r ¼ �0.09, P ¼.39) or width (r ¼ �0.04, P ¼ .7) and depth (r ¼ �0.01, P ¼ .91) of the LFC. Knee flexion angle was negatively correlatedwith weight (r ¼ �0.44, P < .001) and BMI (r ¼ �0.46, P < .001). Conclusions: Using an accessory medial portal forindependent femoral tunnel drilling, with maximal knee hyperflexion, in ACL reconstruction consistently produced tunnellengths greater than 30 mm with no posterior wall fractures. Tunnel lengths tend to be longer with increasing patientheight, mass, and larger LFC dimensions. Maximum knee flexion angle achieved intraoperatively tends to be less forpatients with increasing weight and BMI. Level of Evidence: Level IV, therapeutic case series.

ne aspect of recent evolution in anterior cruciate

Oligament (ACL) reconstruction surgery has beenemphasis on a more anatomic reconstruction.1 This hasbeen described using both single- and double-bundletechniques.2,3 In either case, the concept is to place theACL graft in a more anatomic, or central, locationwithin the native ACL footprints on both the tibia andfemur. On the femoral side, placing the tunnel more

From the Department of Orthopaedic Surgery (M.T.), University of Min-esota, Minneapolis, Minnesota; Elite Sports Medicine/Connecticut Children’sedical Center (M.D. Milewski), Hartford, Connecticut; Department ofrthopaedic Surgery (E.W.C., S.F.B., J.M.H., M.D. Miller), University ofirginia, Charlottesville, Virginia; and Northwest Iowa Bone, Joint, andports Surgeons (J.C.H.), Spencer, Iowa, U.S.A.The authors report that they have no conflicts of interest in the authorship

nd publication of this article.Received December 1, 2011; accepted August 17, 2012.Address correspondence to Marc Tompkins, M.D., Department of Ortho-

aedic Surgery, University of Minnesota, 2450 Riverside Avenue S, Suite200, Minneapolis, MN 55454, U.S.A. E-mail: tompkinsm@hotmail.com� 2013 by the Arthroscopy Association of North America0749-8063/11802/$36.00http://dx.doi.org/10.1016/j.arthro.2012.08.019

38 Arthroscopy: The Journal of Arthroscopic and Related Su

central in the ACL footprint has resulted in the graftbeing placed lower on the wall of the lateral femoralcondyle within the notch than is typical with a trans-tibial technique using an over-the-top guide; the moreanatomic position also places the graft in a more hori-zontal orientation.4 In contrast to a more vertical graft,studies have found that a more anatomic, or horizontal,graft provides better rotational control in addition toanterior-to-posterior translational stability.5-8 Given thecurrent literature, this is the main advantage of focusingon a more anatomic reconstruction technique.One technique commonly used to place the graft in

a more anatomic position is to drill the femoral tunnelindependently of the tibial tunnel in an antegradefashion using a medial portal.3 Initially this meant usinga traditional anteromedial parapatellar arthroscopyportal.9 The anteromedial portal can be placed in sucha way as to allow access to the femoral ACL footprintand still be useful for other intra-articular procedures.Visualization of the ACL femoral footprint with thearthroscope in the lateral portal, however, can bedifficult. Use of an accessory medial portal for femoral

rgery, Vol 29, No 2 (February), 2013: pp 238-243

Fig 1. View of the outlined anterior cruciate ligament femoralfootprint from the anteromedial portal after shaving of theremnant tissue.

FEMORAL TUNNEL LENGTH IN ACL RECONSTRUCTION 239

tunnel drilling has been described by some authors asa method of separating a femoral tunnel drilling portalfrom a traditional anteromedial parapatellar portal thatcan be used for other intra-articular procedures.10 Inaddition, the accessory medial portal can help to facil-itate better visualization because it allows the surgeonto place the arthroscope in the anteromedial portal.11 Ithas been proven that use of the accessory medial portalconsistently places the tunnel in the center of thefemoral ACL footprint.12

Use of a medial portal for independent femoral tunneldrilling has been criticized on the basis that tunnels willbe too short or that the surgeon risks posterior wallblowout of the lateral femoral condyle.13 To counter-balance that and also provide better exposure to thenotch, the technique has subsequently evolved toemploy hyperflexion of the knee.5 The purpose of thisstudy was to evaluate tunnel length during indepen-dent femoral tunnel drilling using an accessory medialportal with the knee in maximal hyperflexion, and thencorrelate the tunnel length and flexion angle withanthropometric data. The hypothesis was that meantunnel length would exceed 30 mm and no posteriorwall fractures would be found.

MethodsInstitutional review board approval was obtained for

this study. All patients undergoing primary ACL recon-struction by the 3 senior authors during a 1-year period,from November 2010 through October 2011, wereincluded in the study. Grafts used were either hamstringor patellar tendon autografts. Patients undergoing revi-sion ACL reconstruction or ACL reconstruction as part ofa grossly unstable knee caused by multiligamentousinjury were excluded. Patients with other intra-articularpathology such as meniscus tears and chondral damagewere not excluded. A total of 106 patients wereconsecutively enrolled in the study. They included 59male patients and 47 female patients with an averageage of 29.6 � 11.6 years (range, 14 to 60).All patients underwent examination under anesthesia

followed by a diagnostic arthroscopy to confirm an ACLtear prior to ACL reconstruction. All patients under-went the same technique for independent femoraltunnel drilling, similar to the “footprint” technique.14

The ACL was first evaluated with the arthroscope ina traditional anterolateral portal. A shaver was used todebride the ACL stump through a traditional ante-romedial portal. A short cuff of tissue was left intact atthe femoral ACL footprint, for visual evaluation of thesize, shape, and location of the footprint14 (Fig 1). Thearthroscope was then moved to the anteromedial portalto allow direct visualization of the entire footprint. An18-gauge spinal needle was inserted into the joint toidentify an acceptable location for an accessory medialportal that allowed a direct line to the center of the

footprint. The location of the portal was more medialand inferior than that of the traditional anteromedialportal. Generally, this placed the portal 1 to 2 cm medialand 5 mm distal to the standard anteromedial portal;however, the location depended entirely on where thespinal needle showed the most desirable spot to reachthe femoral ACL footprint. Within the joint, care wastaken to ensure that the portal entered just superior tothe medial meniscus. With a view of the entire foot-print, including the posterior extent, an underwaterbovie was used to mark the center of the cuff of tissueand create a small area of exposed bone. A micro-fracture awl was advanced at the site of exposed bonein the center of the footprint. A Beath needle wasintroduced and seated in the hole created by the awl,and then a metal sleeve was placed on the medial andposterior aspect of the needle to protect the medialfemoral condyle (Figs 2 and 3).With the Beath needle left in place, the knee was

maximally hyperflexed (Fig 4). The Beath needle wasthen advanced through the lateral femoral condyle. A4.5-mm EndoButton (Smith & Nephew, Andover, MA)drill bit was introduced over the Beath needle. TheBeath needle and drill bit were removed, the depth ofthe tunnel was measured with a depth gauge, and theBeath needle was reinserted (Fig 5). At this same time,a sterile goniometer was used to measure knee flexionangle. Next, an appropriately sized, partially flutedreamer was introduced over the Beath needle. Thereamer was advanced to within 5 to 6 mm of the lateralcortex, and then the protective metal sleeve wasremoved. For each step the metal sleeve was left inplace and care was taken to protect the medialmeniscus and medial femoral condyle with introduc-tion of each instrument. A loop of suture was thenplaced through the Beath needle and advanced to the

Fig 2. Placement of the Beath pin in the center of the anteriorcruciate ligament femoral footprint.

Fig 4. Intraoperative hyperflexion of the knee to 135�.

240 M. TOMPKINS ET AL.

femoral tunnel aperture for later retrieval through thetibial tunnel once the tibial tunnel was prepared (Fig 6).Finally, the knee was taken out of hyperflexion fortibial tunnel preparation.A secure database was maintained with all patient

information. Tunnel length and knee flexion angle at thetime of drilling were recorded. In addition, height,weight, and body mass index (BMI) were recorded tocorrelate patient demographics with tunnel length andknee flexion angles. The width and anteroposteriorlength of the lateral femoral condyle (LFC), as measuredon magnetic resonance images or plain radiographs,were also correlatedwith tunnel length and knee flexionangles. On magnetic resonance images, the width anddepth of the LFCweremeasured between 15 and 20mm

Fig 3. The medial femoral condyle is protected by a metalsleeve as the reamer enters the knee joint anterior to thesleeve.

proximal to the joint line at the point of greatest widthof the LFC; these were always measured at a right angleto one another (Fig 7). On plain radiographs, the depthof the LFC was measured on the lateral view and thewidth was measured on the anteroposterior view, againapproximately 15 to 20 mm proximal to the joint line;magnification error was taken into account by the PACSsoftware used in themeasurements.Magnetic resonanceimages (77 patients) or plain radiographs (9 patients)were available for only 86 patients. Correlation statisticalanalysis using Spearman’s rho was applied to all vari-ables; P < .05 was felt to be significant.

ResultsThe aggregate data for each variable are summarized in

Table 1. Average tunnel length was 37.0 � 3.3 mm(range, 26 to 45), with all but one tunnel greater than 30mm. Average knee flexion angle was 134.4 � 5.0�

(range, 122� to 147�). AverageBMIwas 27.8�7.8 kg/m2

Fig 5. Depth gauge measurement of femoral tunnel.

Fig 6. Femoral tunnel within anterior cruciate ligamentfootprint; view from accessory medial portal.

Table 1. Aggregate Results for Each Variable

Mean Range

Age 29.6 � 11.6 14-60Gender 59M, 47FHeight (cm) 173.0 � 9.2 195.6-150Weight (kg) 83.4 � 24.3 172.2-48Body mass index 27.8 � 7.8 63.3-17.7Tunnel length (mm) 37.0 � 3.3 45.0-26.0Knee flexion angle 134.4 � 5.0� 147.0-122.0�

Width (mm) 32.9 � 4.3 60.3-26.1Depth (mm) 61.1 � 5.0 73.3-48.1

FEMORAL TUNNEL LENGTH IN ACL RECONSTRUCTION 241

(range, 17.7 to 63.3), calculated from an average heightof 173.0 � 9.2 cm (range, 150 to 195.6) and weightof 83.4 � 24.3 kg (range, 48 to 172.2). Average LFCwidthwas 32.9� 4.3 cm (range, 26.1 to 60.3), and depthwas 61.1 � 5.0 cm (range, 48.1 to 73.3). The correlationresults are summarized in Table 2. Height (r ¼ 0.5, P <.001) and weight (r ¼ 0.33, P ¼ .001), but not BMI (r ¼0.14, P ¼ .17), correlated positively with tunnel length.Width (r¼ 0.46, P< .001) and depth (r¼ 0.38, P< .001)of the LFC also correlated positively with tunnel length.Knee flexion angle was not correlated with tunnel

Fig 7. Width (horizontal gray line) and depth (vertical whiteline) of the lateral femoral condyle (LFC) as measuredbetween 15 and 20 mm proximal to the joint line at the pointof greatest width and full depth of the LFC.

length (r ¼ �0.09, P ¼ .39) or width (r ¼ �0.04, P ¼ .7)and depth (r ¼ �0.01, P ¼ .91) of the LFC. Knee flexionangle was negatively correlated with weight (r ¼ �0.44,P < .001) and BMI (r ¼ �0.46, P < .001). There wasno posterior wall blowout in any tunnel.

DiscussionUse of an accessory medial portal for independent

femoral tunnel drilling with the knee in maximalhyperflexion creates an average tunnel length of 37.0mm regardless of patient size, and all tunnels but one(105/106 patients) were greater than 30 mm. Tworecent studies using an accessory medial portal withaverage femoral tunnel lengths of 35.6 and 34.3 mmare consistent with our data.15,16 The ideal or minimaltunnel length remains unclear, but for nearly all formsof femoral graft fixation, this tunnel should be a longenough tunnel to hold an adequate amount of graft tofoster healing of the graft.17 In addition, there was noposterior wall blowout in any tunnel, including drill bitdiameters as large as 11 mm. Given these findings, ifappropriate technique is used, this approach should beapplicable in primary ACL reconstruction and aid ina more anatomic reconstruction.The biggest concerns about any type of medial portal

drilling have been whether other techniques mightproduce better tunnel lengths, which would then bemore conducive to healing, and whether there is a riskof posterior wall blowout.13,18 On the basis of data fromour study, moving from a traditional anteromedialportal to an accessory medial portal appears to helpaddress these concerns. In the opinion of the authors,one reason for this may be that an accessory medialportal placed as medial as possible causes the trajectoryof the femoral tunnel to be angled more anterior withinthe lateral femoral condyle in the axial plane and,therefore, away from the posterior wall. This, combinedwith hyperflexion of the knee, allows the surgeon todrill a femoral tunnel with less concern for overlyshortened tunnels or posterior wall blowout of thelateral femoral condyle. We feel that the angle of dril-ling produced with this technique is similar to that ina previous study suggesting that 20� to 40� in the axial

Table 2. Correlation Results Using Spearman Rho

Age Gender Height Mass BMITunnelLength

Knee FlexionAngle

Width ofLFC

Depth ofLFC

Age Correlation coefficient 1.0 �0.12 �0.08 0.25y 0.33y �0.28y �0.20* �0.17 �0.001Significancez 0.22 0.43 0.01 0.00 0.00 0.04 0.11 0.99

Gender Correlation coefficient 1.0 �0.04 0.06 0.06 0.11 0.14 0.01 0.03Significancez 0.70 0.56 0.54 0.27 0.15 0.92 0.82

Height Correlation coefficient 1.0 0.45y 0.06 0.50y �0.05 0.67y 0.63ySignificancez 0.00 0.56 0.00 0.61 0.00 0.00

Mass Correlation coefficient 1.0 0.89y 0.33y �0.44y 0.36y 0.34ySignificancez 0.00 0.00 0.00 0.00 0.00

BMI Correlation coefficient 1.0 0.14 �0.46y 0.07 0.07Significancez 0.17 0.00 0.50 0.51

Tunnel length Correlation coefficient 1.0 �0.09 0.46y 0.38ySignificancez 0.39 0.00 0.00

Knee flexion Correlation coefficient 1.0 0.04 0.01angle Significancez 0.70 0.91Width of LFC Correlation coefficient 1.0 0.57y

Significancez 0.00Depth of LFC Correlation coefficient 1.0

SignificancezBMI, body mass index; LFC, lateral femoral condyle.*Correlation is significant at the 0.05 level.yCorrelation is significant at the 0.01 level.zTwo-tailed significance.

242 M. TOMPKINS ET AL.

plane is the best angle for drilling from a medialportal.19

The correlations derived from this study are also infor-mative. Height and weight correlated positively withtunnel length, as did lateral femoral condyle size, whichmakes sense in that a larger person should have longertunnels. Also, weight and BMI correlated negatively withknee flexion angle, which seems intuitive as well giventhatmore soft tissuewill prevent a greater angleofflexion.Fortunately, none of these variables ultimately causestunnels to be too short for adequate graft fixation. Inter-estingly, tunnel length did not correlate with kneeflexionangle. Likely, this is because the range of knee flexionangles (122� to 147�) was too small to produce a correla-tion. The minimum angle necessary to carry out thistechnique is not clear; however, it does appear that asmuch flexion as possible will help to produce tunnels ofadequate length and avoid posterior wall blowout.20

This study shows that the technique of using anaccessory medial portal for independent femoral tunneldrilling in ACL reconstruction is promising. There isa learning curve, however, so surgeons must pay carefulattention while learning this new technique.21-23

In particular, care must be taken to protect the medialfemoral condyle and medial meniscus with introductionof instrumentation through the accessory medial portal;use of a metal sleeve and a partially-fluted drill bit asdescribed in the methods can be helpful.

LimitationsLimitations of this study include the fact that there was

no control group. Because concerns have been raised

about tunnel length using this technique, this studyfocused only on evaluating tunnel length for this tech-nique. This was also a time zero study, and therefore, thelong-term clinical effects will need to be followed.

ConclusionsIn anterior cruciate ligament reconstruction, use of an

accessory medial portal for independent femoral tunneldrilling, with the knee in maximal knee hyperflexion,consistently produced tunnel lengths greater than 30mm with no posterior wall fractures. Tunnel lengthstend to be longer with increasing patient height, mass,and larger lateral femoral condyle dimensions. Themaximum knee flexion angle achieved intraoperativelytends to be smaller for patients with increasing weightand body mass index.

AcknowledgmentThe authors thank the staff at the University of Vir-

ginia Outpatient Surgery Center for their assistance inthis study. In particular, the logistical input and assis-tance from Rebecca Rose, R.N., were critical in thedevelopment of the study.

References1. Kopf S, Musahl V, Tashman S, et al. A systematic review of

the femoral origin and tibial insertion morphology of theACL. Knee Surg Sports Traumatol Arthrosc 2009;17:213-219.

2. Buoncristiani AM, Tjoumakaris FP, Starman JS, et al.Anatomic double-bundle anterior cruciate ligamentreconstruction. Arthroscopy 2006;22:1000-1006.

FEMORAL TUNNEL LENGTH IN ACL RECONSTRUCTION 243

3. Steiner M. Anatomic single-bundle ACL reconstruction.Sports Med Arthrosc 2009;17:247-251.

4. Miller CD, Gerdeman AC, Hart JM, et al. A comparison of2 drilling techniques on the femoral tunnel for anteriorcruciate ligament reconstruction. Arthroscopy 2011;27:372-379.

5. Bedi A, Musahl V, Steuber V, et al. Transtibial versusanteromedial portal reaming in anterior cruciate liga-ment reconstruction: An anatomic and biomechanicalevaluation of surgical technique. Arthroscopy 2011;27:380-390.

6. Steiner ME, Battaglia TC, Heming JF, et al. Independentdrilling outperforms conventional transtibial drilling inanterior cruciate ligament reconstruction. Am J Sports Med2009;37:1912-1919.

7. Sadoghi P, Kropfl A, Jansson V, et al. Impact of tibial andfemoral tunnel position on clinical results after anteriorcruciate ligament reconstruction. Arthroscopy 2011;27:355-364.

8. Zantop T, Diermann N, Schumacher T, et al. Anatomicaland nonanatomical double-bundle anterior cruciateligament reconstruction: Importance of femoral tunnellocation on knee kinematics. Am J Sports Med 2008;36:678-685.

9. Harner CD, Honkamp NJ, Ranawat AS. Anteromedialportal technique for creating the anterior cruciate liga-ment femoral tunnel. Arthroscopy 2008;24:113-115.

10. Pombo MW, Shen W, Fu FH. Anatomic double-bundleanterior cruciate ligament reconstruction: Where are wetoday? Arthroscopy 2008;24:1168-1177.

11. Araujo PH, van Eck CF, Macalena JA, et al. Advances inthe three-portal technique for anatomical single- ordouble-bundle ACL reconstruction. Knee Surg SportsTraumatol Arthrosc 2011;19:1239-1242.

12. Kopf S, Pombo MW, Shen W, et al. The ability of 3different approaches to restore the anatomic anteromedialbundle femoral insertion site during anatomic anteriorcruciate ligament reconstruction. Arthroscopy 2011;27:200-206.

13. Bedi A, Raphael B, Maderazo A, et al. Transtibial versusanteromedial portal drilling for anterior cruciate ligament

reconstruction: A cadaveric study of femoral tunnellength and obliquity. Arthroscopy 2010;26:342-350.

14. Bedi A, Altchek DW. The “footprint” anterior cruciate liga-ment technique: An anatomic approach to anterior cruciateligament reconstruction. Arthroscopy 2009;25:1128-1138.

15. Ilahi OA, Ventura NJ, Qadeer AA. Femoral tunnel length:Accessory anteromedial portal drilling versus transtibialdrilling. Arthroscopy 2012;28:486-491.

16. Wang JH, Kim JG, Ahn JH, et al. Is femoral tunnel lengthcorrelated with the intercondylar notch and femoralcondyle geometry after double-bundle anterior cruciateligament reconstruction using the transportal technique?An in vivo computed tomography analysis. Arthroscopy2012;28:1094-1103.

17. Zantop T, Ferretti M, Bell KM, et al. Effect of tunnel-graftlength on the biomechanics of anterior cruciate ligament-reconstructed knees: Intra-articular study in a goat model.Am J Sports Med 2008;36:2158-2166.

18. Lubowitz JH, Konicek J. Anterior cruciate ligament femoraltunnel length: Cadaveric analysis comparing anteromedialportal versus outside-in technique. Arthroscopy 2010;26:1357-1362.

19. Hamilton SC, Jackson ER 2nd, Karas SG. Anterior cruciateligament femoral tunnel drilling through anteromedialportal: Axial plane drill angle affects tunnel length.Arthroscopy 2011;27:522-525.

20. Gelber PE, Erquicia J,Abat F, et al. Effectiveness of a footprintguide to establish an anatomic femoral tunnel in anteriorcruciate ligament reconstruction: Computed tomographyevaluation in a cadaveric model. Arthroscopy 2011;27:817-824.

21. Hohmann E, Bryant A, Tetsworth K. Tunnel positioningin anterior cruciate ligament reconstruction: How long isthe learning curve? Knee Surg Sports Traumatol Arthrosc2010;18:1576-1582.

22. Lubowitz JH. Anteromedial portal technique for theanterior cruciate ligament femoral socket: Pitfalls andsolutions. Arthroscopy 2009;25:95-101.

23. Sohn DH, Garrett WE Jr. Transitioning to anatomicanterior cruciate ligament graft placement. J Knee Surg2009;22:155-160.