A recent curbside consult following the International Patellofemoral Study Group meeting brought to my mind how the treatment of recurrent patellar instability has evolved and how much further it has to go.
I finished my residency in 1986. This dates me and the treatment of recurrent patellar instability in the United States at that time. Circa 1986, text books heralded the invasive open treatments that range from the Insall proximal tube approach to the biomechanically ill-advised Hauser posterior medialization of the tibial tuberosity. Use of lateral release was spreading across the United States with positive “turn up” sign (patella turned up 90° for those of you fortunately too young to know this term) demonstrating it had been done sufficiently. This was then adapted to an arthroscopic approach, which was thought to be better as it was “minimally invasive.” Another popular approach at the time was to combine lateral release with tibial tuberosity medialization or to add tuberosity medialization when the lateral release failed.
The recent curbside consult was an inquiry regarding a planned lateral release and Elmslie-Trillat for a 19-year-old girl with recurrent lateral patellar instability (RPI) and a tibial tubercle–trochlear groove (TT-TG) of 19 mm. The answer was and is complicated. The first question to understand is: Why is the patient having RPI? What are the risk factors — trochlear dysplasia, patellar alta, femoral anteversion, collagen disorder, age, activity, mechanism of injury, and the now popular TT-TG measurement of tuberosity position? Once those factors are known, how do we apply them to improve our decision making?
In most RPI cases, we thought the common denominator was patholaxity of the medial patellofemoral ligament (MPFL). The work of the Rizzoli institute on the medial patellotibial ligament (MPTL) and that of John P. Fulkerson, MD, and Miho J. Tanaka, MD, on the variation of “MPFL” attachment to the medial distal quadriceps tendon in addition to the patella now point to a potentially improved vernacular: the medial patellofemoral complex (MPFC).
Regardless of the name, once these medial tethering structures are compromised, the next restraint is the lateral retinaculum. Thus, it no longer makes biomechanical sense to treat recurrent patellar lateral instability with isolated lateral release. Yes, in a subset of patients, the lateral structures may, indeed, be too restraining/tight. In these cases, a lateral lengthening now popularized by several surgeons including Roland M. Biedert, MD, allows balancing of the lateral structures while maintaining control.
Treatment options evolve
Use of Elmslie-Trillat medial tibial tubercle transfer was effective as seen in the literature for the treatment of RPI, but that does not mean it is the optimal treatment. As we know, the Hauser posteromedial tibial tubercle transfer was also effective, yet it eventually led to patellofemoral (PF) arthritis. In the 1980s, the lead surgeon often said the amount of tuberosity medialization was “just enough” and often referred to as moving the tuberosity to where it “wants to be with the knee at 90° flexion,” that is, in line with the trochlear groove. We now know (well, we think we know) the tuberosity is approximately 10 mm to 13 mm lateral to the midline of the trochlear groove in non-dislocators (i.e., normal knees). Ryosuke Kuroda, MD, PhD, and Jack T. Andrish, MD, showed the potential (deleterious) overload of the medial PF compartment with over-medialization of the tuberosity. David H. Dejour, MD, and others showed the value of attempting to quantitate where the tuberosity is, i.e., how can you move it without knowing where it is?
In the early 1990s, it was rare for a U.S. surgeon to know about the TT-TG or use it in surgical planning. Gradually in the 2000s, all residents and fellows learned the term, how to measure it and how to use it in surgical planning. Thus, a TT-TG of 19 mm would suggest moving the tuberosity medially to achieve a TT-TG of 10 mm to 13 mm. This was the first objective measurement to be used in surgical planning. However, as previously noted, the accepted common pathology to all RPI is patholaxity of the MPFC — not the tuberosity position. Spike Erasmus, MD, and Thomas E. Fithian, MD, both have large series of successful treatment of RPI with MPFL reconstruction without tuberosity medialization. Thus, the precise role of the position of the tuberosity in RPI was and is being questioned by these surgeons and others.
Then, papers re-evaluating the TT-TG began appearing. Diane L. Dahm, MD, and colleagues at Mayo Clinic showed the discrepancy between the TT-TG as measured by CT and MRI. Examining why this occurred, investigators looked at the role of knee flexion which was shown to decrease the TT-TG. Gerd Seitlinger, MD, published a means to eliminate the role of “other factors” in altering the TT-TG measure of the tuberosity position. He referenced the tuberosity to the PCL (TT-PCL). This highlighted the difficulty in even measuring a TG in a patient with trochlear dysplasia, and the significance of when there is rotation of the femur (anteversion), as well as knee femoral tibial rotation.
We published confirmatory data and more recently Keiji Tensho, MD, emphasized the role of rotation of the tibia relative to the femur in altering the TT-TG compared with TT-PCL. Furthermore, Robert A. Teitge, MD, and Leon E. Paulos, MD, have long emphasized the need to evaluate all axial plane contributions to altered patellar tracking and the importance of treating the pathology. That is, if the problem is femoral anteversion, they would advocate femoral derotation and tibial derotation for tibial torsion. Nevertheless, many surgeons consider this approach excessive with an imbalance of risk vs. reward. Fotios P. Tjoumakaris, MD, James P. Bradley, MD; Fulkerson and others continue to report excellent outcomes of anteromedial tibial tubercle transfer; yet this procedure is now used primarily to unload lateral and distal patellar chondrosis in selected patients and not as an isolated first-line treatment of RPI.
As previously noted, the common denominator in patients with RPI is MPFC patholaxity. In the 1990s, most U.S. surgeons were not aware of the specific medial retinacular (MPFC) anatomy and rather, the medial sleeve was imbricated, reefed or advanced. Obviously, these surgeries would include the MPFC with its MPFL and medial quadriceps tendon femoral ligament (MQTFL) components and shortening/tightening of it, but now we can now see in most cases it would not be anatomically based and could contribute to abnormal PF forces. In the 2000s, U.S. surgeons began to adopt MPFL reconstruction and rapidly it became commonly accepted and practiced. Like many new techniques with rapid clinical adoption, MPFL reconstruction was followed by a multitude of complications.
To avoid complications, the relevant anatomy and biomechanics must be identified and restored. The attachment to the patellar and distal quadriceps is highly variable as shown by Fulkerson and Tanaka. There is not one point of attachment. Instead, it is a relatively broad area. In their cadaveric studies, the mean attachment point is just proximal to the articular surface of the medial patella (Tanaka’s point, as coined by Fulkerson), which confirms the rationale for Fulkerson’s MQTFL reconstruction. However, the effect on attachment positioning for the MPFC appears to be much less sensitive (with less effect on the length changes between the attachment sites) than the femoral side and contributing to variability in the success of various patellar attachment techniques for many years. Nevertheless, as a potential devastating effect of patellar tunnels is patellar fracture, the soft tissue fixation options (distal medial quadriceps of Fulkerson MQTFL and possibly soft tissue only fixation at the patella, such as the fascial tunnel of Mikashima) of the MPFC are worth further consideration.
On the femoral side, there was a rush to make finding the attachment point simple. From nine cadavers, Philip Schoettle, MD, PhD, gave us a point, which was the mean of a “cloud of points” and in a similar manner, Joanna M. Stephen, PhD, and Andrew A. Amis, FREng, DSc, reported a somewhat similar point derived from a cloud of points. Vicente Sanchis-Alfonso, MD, PhD, and others have shown that although “the Schoettle or Stephen” point is useful in determining the region of the attachment, it is not the attachment site for every patient. Recreating the anatomy remains important, and the readily identifiable adductor tubercle is a useful starting reference. The take-home message is, place the first point just distal to the tubercle.
The key though is to not stop here, but rather to test the length changes between the attachment points during knee range of motion. The femoral point is then adjusted until the proposed final graft (note the drill pin is in the center of the final socket, and the graft in some tenodesis techniques will be eccentric to this) becomes lax as the knee is flexed. Although debated to this date, the distance between the patellar and femoral attachments is greatest at +/-20° flexion. This is my rational for fixing the graft at this degree of flexion with the graft “out to length;” that is, there is no tension in the graft. It is still possible after final fixation for the patella to displace one or two quadrants. The new graft acts as a tether with tension generated in the graft only with lateral displacement. Finally, as the graft does not see tension in full range of motion and prospective patients are at risk of stiffness, early (protected) motion is essential.
So much for a 60-second curbside consult, and I have not even discussed trochlear dysplasia or patellar alta. There is so much to learn and explore. This is not a static field, and much of what I have suggested may not stand the test of time.
Disclosure: Farr reports he receives royalties from DJO Global and is a patent holder for DJO Global.
Fulkerson JP, et al. Arthrosc Tech. 2013; doi: 10.1016/j.eats.2013.01.002.
Kuroda R, et al. Articular cartilage contract pressure after tibial tuberosity transfer. A cadaveric study. Am J Sports Med. 2001;29:403-409. PMID: 11476376.
Sanchis-Alfonso V, et al. Knee Surg Sports Traumatol Arthrosc. 2015; doi: 10.1007/s00167-015-3523-x.
Tanaka M, et al. Arthroscopy. 2015; doi: 10.1016/j.arthro.2015.04.028.
Zaffagnini S, et al. Knee Surg Sports Traumatol Arthrosc. 2014; doi: 10.1007/s00167-013-2751-1.