Patellofemoral Update focuses on the causes, prevention and treatment of patellofemoral disorders. The blog is sponsored by The Patellofemoral Foundation whose mission is to improve the care of individuals with anterior knee pain through targeted education and research. The Patellofemoral Foundation offers additional online education resources on its website.

BLOG: Update on medial patellofemoral anatomy, implications for reconstruction

The anatomy of the medial patellofemoral ligament has been widely described as having an origin on the medial femur and a “sail-shaped” or fan-shaped attachment on the patella. Understanding the anatomy of the medial patellofemoral ligament allows us to recreate the attachment points and isometric function of the ligament during reconstruction in the treatment of patellar instability. Although many studies have described the ideal location for femoral tunnel placement, few have attempted to identify the ideal point of fixation to recreate the patellar attachment.

Many consider the appropriate patellar fixation point of a single-strand graft to be in the proximal half, or at the junction of the proximal and middle thirds of the patella, although few studies have supported this with biomechanical evidence. In one anatomical study, Nomura and colleagues described the midpoint of the patellar attachment as 27+/-10% distal to the superior pole of the patella.

The term ‘MPFL’ is not anatomically accurate

Miho J. Tanaka

Recent cadaveric studies have shown the presence of proximal medial patellofemoral ligament (MPFL) fibers that attach to the quadriceps tendon, which can be visualized during dissection from the articular surface after removal of layer 3. Mochizuki described a consistent attachment of the fibers to the vastus intermedius tendon in a dissection study of 16 knees, as did Baldwin in a large series of 50 knees. Placella’s group described that an attachment to the quadriceps tendon was present in seven of 20 cadaveric knees.

John P. Fulkerson, MD, and Cory M. Edgar, MD, PhD, have termed these proximal fibers that attach to the quadriceps tendon as the medial quadriceps tendon femoral ligament (MQTFL), a distinct anatomic bundle of fibers that was proximal to the MPFL and attached to the quadriceps tendon.

In my cadaveric study of 38 knees, I found the fibers attaching to the patella and quadriceps tendon appeared to have a common origin on the femur. The fibers attached to both the patella and quadriceps tendon in all but two knees. In quantifying the distribution of these fibers, I found that 57.3+/-19.5% of these fibers attached to the patella, and the remainder attached to the quadriceps tendon. However, there was a considerable amount of variability in MPFL morphology in this study (range 0% to 100%).

Figure 1. In our study, we described a point (yellow asterisk) at the junction of the line along the medial quadriceps tendon (dotted purple line) and the superior articular border that consistently approximated the MPFC midpoint (blue line).

Source: Tanaka MJ

The presence of fibers attaching to both the patella and quadriceps tendon suggests the term “MPFL” is not completely accurate. Despite the varied attachment points, it appears the MPFL/MQTFL has a single origin and serves as a one complex to stabilize the patella. Because the term “MPFL” implies an anatomic attachment to the patella, Fulkerson and I began using the term medial patellofemoral complex (MPFC) to refer to these fibers and to describe the stabilizing forces of this ligament on the patellofemoral joint, without limiting the attachment site to a given anatomic location.

Should we be utilizing double-bundled grafts?

Because of the wider attachment on the patella and quadriceps tendon than its origin on the femur, several authors have promoted the technique of using a double-bundled graft to recreate the anatomy of this complex. Kang and colleagues described two components of MPFL fibers, using the term “superior-oblique bundle” and “inferior-straight bundle.” While they noted these bundles could not be differentiated as separate anatomic ligaments, they described the angle between the bundles as 15.1+/-2.1°. The clinical significance of this is not yet known, but the authors suggested the bundles may vary in their roles as dynamic versus static stabilizers.

Furthermore, length differences between the two “bundles,” or attachment sites, have been described. Mochizuki and colleagues showed the length of the MPFL fibers from the origin to the medial patella was 56.3+/-5.1 mm vs. 70.7+/-4.5 mm to the quadriceps tendon. Similarly, in my previous study, there was a 7-mm difference in length from the femoral origin to the patella or the quadriceps tendon that was statistically significant. Given the implications for changes in isometry that occur with variable graft position, further biomechanical studies are needed to assess the benefit of such a graft.

Where is the midpoint of the MPFC attachment?

Despite reports on using double-bundled grafts for patellar stabilization, most patellofemoral surgeons are utilizing a single-strand graft with good results. With a wide attachment site on the quadriceps tendon and patella, where should we place the graft? Anatomically, it makes sense to place this at the center of its attachment.

In our most recent study, we sought to identify the midpoint of the MPFC attachment. Using a digital image analysis program on images of cadaveric dissections performed by myself and Fulkerson, we identified the midpoint of the MPFC by computing a bisecting line through the maximal proximal-distal width of its attachment. The results of our study showed two interesting findings:

The midpoint of the MPFC attachment was more proximal than our current guidelines for graft placement, at 2.3+/-15.8% of patellar articular length (PAL) from the superior pole of the patella. In an average 36-mm patella, this would equate to less than 1 mm from the apex of the patellar articular surface.

In addition, we described a reference point (Figure 1) at the junction of the line along the medial quadriceps tendon and the superior articular surface of the patella. We found this point described the midpoint of the MPFC in 64% (16/25) knees, and the remainder were proximal to this by an average of 5.3+/-8.6% PAL (approximately less than 2 mm).

In all knees, the midpoint of the MPFC attachment were at or proximal to the junction of the medial border of the quadriceps tendon and patella. This suggests that midpoint of this complex may actually be proximal to where many of us are currently placing the graft on the patella and could be better reproduced with fixation on the quadriceps tendon. Further biomechanical studies are needed to show the functional, as opposed to a purely anatomical, midpoint of this complex, as well as clinical outcomes studies to determine the optimal site for graft fixation.

MPFC variability and individualized surgery

Throughout the multiple anatomical studies describing the MPFC, the common theme is the variability in the findings and morphology of this complex. In the above study, 22 of 25 knees had attachments to both the quadriceps tendon and patella; two had attachments to the quadriceps tendon only; and one attached entirely to the patella.

This variability exists within paired knees of individuals, as well. In my anatomical study describing the variability of these findings, I reported on the differences in MPFL morphology between the right and left knees in 14 sets of paired cadaveric knees. Two pairs had identifiable MPFL fibers in one knee, but no MPFL in the contralateral knee. Another had no visible MPFL fibers in either knee. In the remaining knees, there was a 24.4+/21.2% side-to-side difference in the percentage of fibers that attached to the patella (vs. quadriceps tendon).

This variability in MPFC anatomy poses difficulty in determining optimal treatments for patellar instability that is not unlike the multifactorial nature of this disorder. Our described point (at the junction of the medial quadriceps tendon and articular border of the patella) consistently re-approximated this midpoint, despite the morphological variability in the appearance of the complex, and may serve as a future guide in anatomic graft placement. Being able to identify each individual’s native anatomy and understand whether to reconstruct or correct this based on the context of their symptoms is an important goal in future research on patellofemoral stabilization surgery.


Baldwin JL. Am J Sports Med. 2009;doi:10.1177/0363546509339909.

Fulkerson JP, et al. Arthrosc Tech. 2013;doi:10.1016/j.eats.2013.01.002.

Kang HJ, et al. Knee Surg Sports Traumatol Arthrosc. 2010;doi:10.1007/s00167-010-1090-8.

Mochizuki T, et al. Knee Surg Sports Traumatol Arthrosc. 2013;doi:10.1007/s00167-012-1993-7.

Nomura E, et al. Knee Surg Sports Traumatol Arthrosc. 2005;doi:10.1007/s00167-004-0607-4.

Tanaka MJ, et al. J Arthroscopy. 2015;doi:10.1016/j.arthro.2015.04.028.

Tanaka MJ, et al. The anatomic midpoint of the attachment of the medial patellofemoral complex. Submitted to Bone Joint J.


Miho J. Tanaka, MD, is assistant professor of orthopedic surgery and director of the Women’s Sports Medicine Program at The Johns Hopkins Hospital in Baltimore.

Disclosure: Tanaka reports no relevant financial disclosures.

See more from Patellofemoral Update