Abstract
Background
The tibial tubercle-trochlear groove measurement (TT-TG), which measures the lateral offset of the tibial tubercle relative to the trochlear groove of the femur, has been utilized as an intraoperative tool to help establish maximum patellofemoral congruency in patients who suffer from patellar instability. We have previously published our approach of establishing how far to transfer the tibial tubercle using intraoperative femoral nerve stimulation in order to achieve congruency from 0-30° of flexion. The technique and clinical outcomes have previously been published in this journal and elsewhere. Here we describe the use of the TT-TG distance to determine how far to transfer the tibial tubercle to achieve our goals and have found that it varies according to the clinical exam features.
Purpose
We intended to determine the effectiveness of using the preoperatively established TT-TG to predict the degree of intraoperative me- dialization of the tibial tubercle to achieve our goal of establishing dynamic congruency of the patella in the trochlear groove when using the previously described femoral nerve stimulation method of estimating dynamic tracking of the patella.
Methods
From the study group of patients used in other publications, we examined 20 knees in 18 patients who had a history of recurrent lateral dislocations and underwent a Fulkerson tibial tubercle transfer. Each knee was dynamically assessed preoperatively by obtaining an MRI at 30° of flexion and complete hyperextension while voluntarily contracting their quadriceps. These were then compared to the intraoperative transfer of the tibial tubercle required to achieve maximum congruency when the femoral nerve was stimulated. We then looked at the preoperative TT-TG measurement to determine its role in predicting what was required at achieving congruency in the context of the quad active MRI findings.
Results
Thirteen knees preoperatively demonstrated a positive J-sign defined as the patella subluxated greater than or equal to 5mm lateral in full extension compared to 30° of flexion. In these patients, the TT-TG was accurate if the distance medialized was 1:1 with the measured TT-TG. In 7 out of the 20 knees, the patella demonstrated a false negative J-sign where the patella was radio- graphically subluxated at 30° of flexion as well as at hyperextension. In this group, the TT-TG underestimated the transfer required for congruency on average 5mm even when using the 1:1 ratio.
Conclusions
The preoperative use of the J-sign is of value when determining the role of the TT-TG measurement and estimating the distance required to intraoperatively achieve congruency when using the femoral nerve stimulation technique. Those that demonstrated a positive J-sign of 5mm or greater, a 1:1 ratio of TT-TG to medialization is most reliable at establishing congruency of the patellofemoral joint. Whereas, those that demonstrated a false negative J-sign even the 1:1 ratio remains inadequate at producing congruency and more medialization is required.
Level of evidence
Level III, Retrospective Observational/Comparative Study
Introduction
Patellar maltracking and instability are largely thought to be the result of an imbalance between bone morphology, soft tissue, and muscular action. In order to surgically correct patellar instability and recurrent dislocations there are numerous options including soft tissue balancing, tibial tubercle transfer, trochleoplasty, as well as rotational osteotomy of the femur1-4. At our institution, the most common procedure performed to treat severe maltracking is an anteromedialization of the tibial tubercle to align the extensor mechanism of the knee thus allowing the patella to glide smoothly within the confines of the trochlear groove. The senior author has previously employed the use of femoral nerve stimulation to intraoperatively estimate the amount of correction required to achieve complete congruency of the patella tracking in the trochlear boundaries of the femur from 0° to 30° of flexion5-8.
Currently, as a preoperative assessment of how far to transfer the tibial tubercle intraoperatively, most surgeons order a preoperative CT scan or an MRI to determine the lateral offset of the tibial tubercle relative to the trochlear groove (TT-TG distance)1,9. Intraoperatively, we consider the TT-TG distance to help guide placement of the tibial tubercle, even though our final placement of the tibial tubercle is based on maximal patellofemoral congruency during active quadriceps extension when the femoral nerve is stimulated. It is our goal to reach complete congruency in this dynamic condition.
The purpose of our study was to determine the effectiveness of using the preoperatively established TT-TG to predict the degree of intraoperative medialization of the tibial tubercle to achieve our goal of establishing dynamic congruency of the patella in the trochlear groove when using the previously described femoral nerve stimulation method of estimating dynamic tracking of the patella.
The goal of our analysis included: 1) To compare our preoperative TT-TG calculations to the actual medializa- tion required for congruency when nerve stimulation of the quadriceps was employed between 30° of flexion and hyperextension and 2) To establish the role of the J-sign in the application of the TT-TG value.
Our hypothesis is that the preoperative quad active MRI-based J-sign assessment plays a significant role in the use of the TT-TG measurement for our purposes of establishing intraoperative femoral nerve-based congruency of the patellofemoral joint.
Materials and Methods
Imaging
We retrospectively identified 20 knees in 18 patients who had a history of recurrent lateral dislocations and obvious patellofemoral maltracking who had undergone Fulkerson tibial tubercle medializations between 2008 and 2011. The average age in the patient population was 23 years with a range of 13-41 years, 11 of which were females and 9 were males. In order to be considered for the study, the patients had to provide consent and must have had all MRI scans prior to surgery that were required for the study (quad active 30° of flexion and full extension as well as full length axial MRI scans from hip to ankle).
Preoperatively, the patients were examined clinically for congruency of the patella in relation to the trochlea of the femur at 30° of flexion and complete hyperextension of the knee. This was objectively quantified by obtaining an MRI with the quadriceps contracted and the leg held at 30° of flexion and at complete extension and measuring the relationship of the lateral aspect of the patella to the lateral aspect of the trochlea of the femur, known as the lateral patellar edge (LPE). Duchman and colleagues described the LPE as the most reliable measurement for quantifying the clinical finding of the J-sign5.
We also obtained a full length lower extremity scan from the hip down to the ankle joint. With this scan, the femoral torsion, tibial torsion, and overall alignment of the lower extremity were measured to look for other potential causes of maltracking.
Measurements
TT-TG: The TT-TG measurement measures the lateral offset of the tibial tubercle relative to the trochlear groove of the femur. An MRI scan was done of the entire lower extremity from the hip to the ankle joint with the leg lying on the table passively positioned in full extension. A reference line was drawn on the posterior aspect of the condyles. A second line was drawn perpendicular to the reference line and through the deepest part of the trochlear groove. A third line was drawn parallel to the second line but through the most anterior aspect of the tibial tubercle. The distance between the second line and the third line along the reference line is the measured TT-TG. Our TT-TG was accurately measured according to Dejour et al. but instead of using CT as they described we used axial MRI images1,9. Schoettle et al. demonstrated that there is no difference between CT and MRI in determining TT-TG distance accurately10.
J-Sign: The J-sign is a clinical finding that describes the route the patella takes as the knee goes from 30° of flexion to full extension during quadriceps contraction. As described by Duchman et al., we quantified the clinical finding of the J-sign based off of the preoperative quad active MRI images that were taken with the quadriceps isometrically contracted and the leg held at full extension and 30° of flexion5. First, a horizontal reference line was drawn along the most posterior aspect of the condyles. A second line was drawn perpendicular to the posterior condyle reference line intersecting the most superior aspect of the lateral condyle. A third line was drawn parallel to the second line; however, instead of intersecting the outer edge of the trochlear groove, this line touches the lateral aspect of the patella. The distance between these lines is the lateral patellar edge (LPE) distance. We modified the LPE slightly by taking into consideration the patellar tilt. This modified LPE measurement as depicted in Figures 1 and 2 was done at the 30° flexion angle and full extension to assess the lateral movement of the patella as the leg moves from 30° of flexion to full extension.
Figure 1. shows a positive J-sign where the patella is subluxated at full extension (A) but within the confines of the trochlear groove when in 30° of flexion (B). This was the characteristic used to define knees 8-20 in table 1. In general, the 1:1 ratio accurately represented the intraoperative medialization in this group with only an average of 0.2mm difference between preoperative and intraoperative measurements.
Figure 2. shows a subluxated patella throughout flexion and extension. the patella is subluxated at full extension (A) and also subluxated in 30° of flexion (B). this was the characteristic used to define knees 1-7 in table 1. In general these knees required even further medialization than the 1:1 ratio. On average the preoperatively measured 1:1 ratio underrepresented the intraoperative medialization by 5.7mm.
If the patella was subluxated by 5mm or more in full extension compared to 30° of flexion, then those were categorized as a “positive” J-sign as shown in Figure 1. On the other hand, there were others that were subluxated throughout 30° of flexion and full extension as shown in Figure 2. In these knees, the patella never medializes as the leg goes from full extension to 30° of flexion. On clinical exam of the J-sign, this would be found to be a “false negative” J-sign. In this study, this group was classified based on if the patella was displaced laterally by less than 5mm in full extension compared to 30° of flexion. For the purpose of our study, rather than being concerned with the congruency of the patella with the trochlea at 30° of flexion and full extension, we concentrated on the trajectory of the patella from 30° of flexion to full extension so that we could correlate the clinical exam features of the J-sign. Table 1 shows the measured LPE in full extension, 30° of flexion, and whether the knee demonstrated a positive or false negative J-sign.
Table 1.
shows the lateral patellar edge (LPE) distance during full extension and in 30° flexion for each knee. The second to last column is the difference between column 2 and 3 and is used to define which group each knee belongs; it is either a false negative J-sign (any value 5mm or less) or shows a positive J-sign (any value greater than 5mm).
| Knee Number | Full Ext LPE (mm) | 30deg Flxn LPE (mm) | Full Ext - 30deg Flxn LPE (mm) | Postive or False Negative J-sign |
|---|---|---|---|---|
| 1 | 5 | 1 | 4 | False Negative |
| 2 | 24 | 20 | 4 | False Negative |
| 3 | 20 | 16 | 4 | False Negative |
| 4 | 18 | 18 | 0 | False Negative |
| 5 | 24 | 23 | 1 | False Negative |
| 6 | 22 | 23 | -1 | False Negative |
| 7 | 7 | 5 | 2 | False Negative |
| 8 | 19 | 6 | 13 | Positive |
| 9 | 23 | 0 | 23 | Positive |
| 10 | 23 | 10 | 13 | Positive |
| 11 | 23 | 6 | 17 | Positive |
| 12 | 30 | 23 | 7 | Positive |
| 13 | 15 | 9 | 6 | Positive |
| 14 | 30 | 10 | 20 | Positive |
| 15 | 24 | 7 | 17 | Positive |
| 16 | 21 | 10 | 11 | Positive |
| 17 | 16 | 4 | 12 | Positive |
| 18 | 15 | 7 | 8 | Positive |
| 19 | 23 | 14 | 9 | Positive |
| 20 | 17 | 8 | 9 | Positive |
Surgical Technique
As described by Laver y et al., prior to surgery all of the patients received a stimulating femoral catheter which was placed near the femoral nerve at the location of the inguinal crease7. With the patient awake, adequate placement was determined by observing that complete quadriceps contraction had been achieved through stimulation of the nerve. Under anesthesia, a preliminary assessment of patellar tracking was again made by stimulation of the nerve. A medial parapatellar approach was made to expose the tibial tubercle and patella so that tracking of the patella could be directly assessed. In order to allow for a “free body” analysis of the distance required to obtain congruency with quadriceps activated, the medial patellofemoral ligament (MPFL) and lateral patellofemoral ligament (LPFL) were severed from the patella and an anteromedialization of the tibial tubercle was performed. The tubercle was then transferred according to three measurements: 1) a 2:1 TT-TG to medialization ratio, 2) a 1:1 ratio of the TT-TG measurement, and finally 3) the point of maximum congruency during active quadriceps contraction through stimulation of the nerve and direct observation of the relationship of the patella to the trochlear edges throughout the 30° of flexion to complete extension motion. The 2:1 and 1:1 TT-TG to medialization ratios are illustrated in Figure 3. After each of the three tubercle transfer distances, the tubercle was temporarily pinned with K wires and held while the femoral nerve was being stimulated. Congruency was assessed at each of the transfer distances by feeling for the patella on the lateral edge of the condyle and under direct visualization of the patella tracking through 30° of flexion to full extension during quadriceps contraction. If the patella continued to displace laterally during the nerve stimulation then the tubercle was transferred more medially. If the patella began to displace medially then the tubercle was transferred to a more lateral position. Once maximum congruency was established, the tubercle was stabilized with 2-3 bicorti- cal screws and the distance of transfer was measured. In order to withstand the forces involved in quadriceps contraction, 2-3 fixation screws and a proximal dovetail shelf were required4.
Figure 3. illustrates the measured TT-TG distance of 20mm (left), the 1:1 TT-TG to medialization ratio so medialization is 20mm or the same distance as the measured TT-TG (middle), and the 2:1 TT-TG to medialization value so medialization is 10mm or half the distance of the measured TT-TG (right).
Results
In 13 out of the 20 knees, the patella was subluxated at full extension, but retracted toward the trochlear groove at 30° of flexion with the quadriceps contracted before and during surgery demonstrating a positive J-sign. In this group, the TT-TG was accurate at predicting the medialization if the distance medialized was a 1:1 ratio. In using the 2:1 ratio as the transfer distance, 100% of the time it did not correct or eliminate the J-sign. The success rate of achieving complete intraoperative dynamic congruency was increased from 0% to 92% when a 1:1
ratio was used when accepting less than a 5mm error. Table 2 shows the measured TT-TG for each knee, the 2:1 ratio, the 1:1 ratio, and the actual distance that the tubercle was transferred at surgery when obtaining maximum congruency during quadriceps contraction.
Table 2.
shows the preoperative measured TT-TG for all 20 knees, the 2:1 ratio predicted medialization value, the 1:1 ratio predicted medialization value, and the actual value medialized when finding maximum congruency intraoperatively during femoral nerve stimulation. The values in parentheses correspond to the amount in which the prediction was off of the intraoperative medialized value (negative=underestimates, positive=overestimates). Knees 1-7 demonstrate the false negative J-sign, whereas knees 8-20 are knees that show a positive J-sign. The 2:1 ratio underrepresented the medialization on average by -12.8mm, whereas the 1:1 ratio underrepresented the medialization on average by 1.9mm. Overall, trends show that the 1:1 ratio is best when patients show a positive J-sign preoperatively (knees 8-12); however, in patients that show a false negative J-sign even the 1:1 ratio consistently underrepresents the actual medialization (knees 1-7).
| Knee Number | Measured TT-TG (mm) | 2:1 Ratio (mm) | 1:1 Ratio (mm) | Nerve stimulation (mm) |
|---|---|---|---|---|
| 1 | 20 | 10 (-18) | 20 (-8) | 28 |
| 2 | 24 | 12 (-12) | 24 (0) | 24 |
| 3 | 28 | 14 (-16) | 28 (-2) | 30 |
| 4 | 18 | 9 (-16) | 18 (-7) | 25 |
| 5 | 23 | 11.5 (-16.5) | 23 (-5) | 28 |
| 6 | 24 | 12 (-18) | 24 (-6) | 30 |
| 7 | 14 | 7 (-19) | 14 (-12) | 26 |
| 8 | 26 | 13 (-15) | 26 (-2) | 28 |
| 9 | 18 | 9 (-11) | 18 (-2) | 20 |
| 10 | 18 | 9 (-6) | 18 (3) | 15 |
| 11 | 17 | 8.5 (-11.5) | 17 (-3) | 20 |
| 12 | 29 | 14.5 (-15.5) | 29 (-1) | 30 |
| 13 | 21 | 10.5 (-9.5) | 21 (1) | 20 |
| 14 | 27 | 13.5 (-11.5) | 27 (2) | 25 |
| 15 | 19 | 9.5 (-10.5) | 19 (-1) | 20 |
| 16 | 21 | 10.5 (-7.5) | 21 (3) | 18 |
| 17 | 24 | 12 (-5) | 24 (7) | 17 |
| 18 | 11 | 5.5 (-8.5) | 11 (-3) | 14 |
| 19 | 30 | 15 (-15) | 30 (0) | 30 |
| 20 | 23 | 11.5 (-13.5) | 23 (-2) | 25 |
In the other 7 knees, the patella was subluxated at both 30° of flexion and in complete extension demonstrating a false negative J-sign. In this group, the TT-TG underestimated the transfer required for congruency at 0 and 30° of flexion on average 5mm with a range of 0mm to 12mm even when using the 1:1 ratio. This group required even further medialization of the tubercle to establish congruency using our method.
Discussion
One of the weakest points of any corrective surgery for patellofemoral instability related to maltracking is how much correction is needed. The senior author has developed an approach based on the dynamics of the extensor mechanism of the knee in establishing congruency of the patella tracking within the boundaries of the femoral trochlea as the knee traverses from extension to early flexion5-8. The distance required for our intraoperative goal is based on the TT-TG measurement; however, it has been our experience that this measurement is not always sufficient for our purposes. We utilized the preoperatively measured TT-TG and modified LPE to determine the amount of medializa- tion required at surgery. It is our belief that this gives consideration to the bony anatomy as well as muscular action. In addition, we give consideration to these factors during surgery by utilizing intraoperative femoral nerve stimulation to activate the quad muscle with the patient under anesthesia. We believe that by considering the active properties of the joint that the tibial tubercle transfer placement is better than if we considered only passive properties of the joint without femoral nerve stimulation. By examining the dynamic properties of the joint both preoperatively and intraoperatively, it is reasonable to think that the tibial tubercle placement would be best at maximizing patellofemoral congruency in the postoperative knee as indicated in the articles by Duchman et al. and Mellecker et al5,8.
In this study, the transfer of the tubercle is greater than is commonly performed; however, the intent of this article is solely to describe the use of the TT-TG to achieve our goal of dynamic congruency. Given the TT-TG values for the normal population, Dejour et al. recommends that the goal of tibial tubercle transfer is to reduce the TT-TG to between 10 to 15 mm1,9. In our study, we found that this was not adequate to achieve complete congruency with femoral nerve stimulation. Our findings suggest in the population with a positive J-sign of greater than or equal to 5mm measured by quad active MRIs that one should medialize the tubercle the entire measured TT-TG (1:1 ratio). If the J-sign measurement is < 5mm then even the 1:1 ratio was inadequate and an additional medialization to achieve complete congruency was needed.
From our results, we made the conclusion that in the false negative J-sign group, the 1:1 TT-TG to medi- alization ratio remains inadequate to reliably produce congruency when the muscle is stimulated. We tried to hypothesize what makes the patella track in this way, and why when it does that it requires further medializa- tion than when the patella shows a positive J-sign. It is our belief that the false negative J-sign group suffers from more severe maltracking because the patella does not appear to engage into the confines of the groove. Given the more severe tracking, the requirement for further medialization is needed to realign the extensor mechanism of the knee to maximize congruency. It was to no surprise that we did not find any specific bony anatomic differences between those that demonstrated a positive J-sign or false negative J-sign. The only conclusion we could make was that those that demonstrated a false negative J-sign needed more medialization of the tubercle. Interestingly, when reading through the intraoperative notes on each of the knees, in the false negative J-sign group, notes were made by the attending surgeon that many of these knees had an oblique takeoff angle of the quadriceps tendon compared to those that showed a positive J-sign. We understand that this was mentioned subjectively and that no objective data was obtained in regards to this finding, but it is a future goal of ours to analyze the takeoff angle of the quadriceps tendon with respects to the femur to objectively verify these intraoperative findings.
Conclusion
Given our goal of achieving complete congruency, we made three general conclusions: 1) The 2:1 ratio of TT- TG to medialization is always inadequate regardless of the group to create maximum patellofemoral congruency at both 30° of flexion and extension with the quadriceps contracted, 2) In the group that demonstrated a positive J-sign, a 1:1 ratio of TT-TG to medialization is more reliable, and 3) In the population of patients where the patella remains subluxated at 30° of flexion and full extension (false negative J-sign), even the 1:1 ratio remains inadequate to reliably produce congruency when the muscle is stimulated. Typically, these knees required more significant medialization than the 1:1 ratio.
Acknowledgments
We thank Dr. Daniel Thedens and Dr. D. Lee Bennett in their help with the MRI analysis and thoughts for the project.
References
- 1.Dejour D, Coultre B. Osteotomies in patellofemoral instabilities. Sports Med Arthrosc Rev. 2007;15:39–46. doi: 10.1097/JSA.0b013e31803035ae. [DOI] [PubMed] [Google Scholar]
- 2.Fulkerson JP. Anteromedialization of the tibial tuberosity for patellofemoral malalignment. Clinical Orthopaedic Related Research. 1983;177:176–181. [PubMed] [Google Scholar]
- 3.Redziniak DE, Diduch DR, Mihalko WM, Fulkerson JP, Novicoff WM, Sheibani-Rad S, Saleh KJ. Patellar instability. The Journal of Bone and Joint Surgery. 2009;91:2264–2275. [PubMed] [Google Scholar]
- 4.Southwick WO, Becker GE, Albright JP. Dovetail patellar tendon transfer for recurrent dislocating patella. JAMA. 1968;204(8):665–669. [PubMed] [Google Scholar]
- 5.Duchman K, Mellecker C, El-Hattab AY, Albright JP. Quantitative MRI of tibial tubercle transfer during active quadriceps contraction. Clinical Orthopaedic Related Research. 2011;469:294–299. doi: 10.1007/s11999-010-1598-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ebinger TP, Boezaart A, Albright JP. Modifactions of the Fulkerson osteotomy: a pilot study assessment of a novel technique of dynamic intraoperative determination of the adequacy of tubercle transfer. The Iowa Orthopaedic Journal. 2007;27:61–64. [PMC free article] [PubMed] [Google Scholar]
- 7.Lavery M, Bell J, Rickelman T, Boezaart A, Albright JP. Patellofemoral realignment: Dynamic intraoperative assessment. The Iowa Orthopaedic Journal. 2005;25:160–163. [PMC free article] [PubMed] [Google Scholar]
- 8.Mellecker C, Ebinger T, Butler P, Albright JP. Southwick-Fulkerson Osteotomy with intraoperative femoral nerve guidance. The Iowa Orthopaedic Journal. 2013;33:XXX. [PMC free article] [PubMed] [Google Scholar]
- 9.Caton JH, Dejour D. Tibial tubercle osteotomy in patella-femoral instability and in patellar height abnormality. International Orthopaedics. 2010;34(2):305–309. doi: 10.1007/s00264-009-0929-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Schoettle PB, Zanetti M, Seifert B, Pfirmann CA, Fucentese SF, Romero J. The tibial tuberosity-trochlear groove distance; a comparative study between CT and MRI scanning. Knee. 2006;13:25–31. doi: 10.1016/j.knee.2005.06.003. [DOI] [PubMed] [Google Scholar]