Medial patellofemoral ligament

Function

Where it runs

The medial patellofemoral ligament (MPFL) is a ligament, a tough band of tissue that holds bones together at a joint¹. It runs from the inner side of the lower femur (thigh bone) to the inner edge of the patella (kneecap)¹.

A systematic review (a study that searches for the research on a question in a set, checkable way) of 67 anatomy studies found these features¹:

Thigh-bone end: a triangle between three bony bumps, the adductor tubercle, the medial epicondyle and the gastrocnemius tubercle¹.
Width there: an area rather than a single point, 9 to 17 mm wide¹.
Kneecap end: 20 to 30 mm wide, mostly on the upper part of the inner border¹.
Shape: an hourglass that fans out at both ends, 56 mm long on average¹.
Thickness: 1.0 mm in men, 0.9 mm in women, in one cited MRI (a scan that shows soft tissue) study¹.

Grey 3D drawing of the bones of a bent knee seen from the inner side, with the thigh bone above, the shin bone below and the kneecap at the front on the right. A pale band labelled MPFL runs forward from a dotted triangle on the inner side of the lower thigh bone to a dotted oval on the inner edge of the kneecap. Leader lines label the bony landmarks around the triangle (AT, AMT, mGT, GT, sMCL, MFE), and arrows at the bottom mark superior, inferior, anterior and posterior. — Where the medial patellofemoral ligament attaches, seen from the inner side of a bent right kneeAframian A, Smith TO, Tennent TD, Cobb JP, Hing CB. Origin and insertion of the medial patellofemoral ligament: a systematic review of anatomy. Knee Surg Sports Traumatol Arthrosc 2017, Figure 2. © The Author(s) 2016. CC BY 4.0. (source, CC BY 4.0)

How the reports differ

Studies disagree on the ligament’s size and even its presence¹. Cadaver studies, using bodies donated for study, have found it in 35% to 100% of specimens¹. Yet one dissection of 50 knees found it in each of them⁴. Lengths ran from 45 to 50 mm in eight cadaver knees to about 60 mm in that dissection^4,5.

In the 50-knee dissection, the ligament had two origins⁴. One was a 10.6 mm strip from a bony groove lying between two of the bumps, the medial epicondyle and adductor tubercle⁴. The other was fibres crossing over from the superficial medial collateral ligament⁴.

In 36 child cadavers aged 2 to 11, the thigh-bone end lay below the growth plate in 23 and at or above it in 13⁶. The growth plate is a layer of cartilage near the end of a growing bone. On average, the attachment sat 3.0 mm below it⁶.

What it holds back

In nine cadaver knees bent to 20 degrees, the MPFL gave 60% of the force holding the kneecap back². Other cadaver studies also found it the main inner restraint, with shares that varied^5,7,8:

Study	Knee bend	MPFL share of the inner restraint
9 cadaver knees	20 degrees	60%²
25 knee specimens	Not given	53% on average⁷
8 cadaver knees	30 degrees	More than 50%⁵
One cadaver study	As the kneecap enters the groove	50% to 60%⁸

In 12 cadaver knees bent from 0 to 90 degrees, it did more of this work than the medial quadriceps tendon-femoral ligament⁹.

Cadaver studies that cut the ligament found these changes^10,11:

In 17 knee specimens, cutting it alone raised outward movement of the kneecap by 50%¹⁰.
In ten specimens, cutting it alone greatly raised outward movement between 20 and 90 degrees of bend¹².
In eight knees, cutting it shifted and tilted the kneecap outward early in the bend¹¹. Pressure fell on the inner side of the kneecap joint and peaked higher on the outer side¹¹.

From a study of six specimens, the authors concluded it restrains mainly when the kneecap is under high outward stress¹³. They judged that it adds little during unloaded knee bending¹³. A systematic review of 55 articles found data suggesting the kneecap sits further outward early in the bend without an MPFL¹⁴. Its authors suggest this plausibly raises the risk of dislocation¹⁴.

How its length and strength change

The 55-article review found the MPFL keeps close to the same length over the first 50 to 60 degrees of bend¹⁴. It then shortens step by step into deep bending¹⁴. In 15 healthy people imaged during a lunge, it was tightest with the knee straight¹⁵. It shortened until about 30 degrees, then stayed near one length¹⁵.

The review found reported failure loads from 72 to 208 newtons (a unit of force)¹⁴. Stiffness, the force needed to stretch it by a given amount, ranged from 8.0 to 42.5 newtons per mm¹⁴. In ten cadaver knees, it failed at 208 newtons on average, after stretching 26 mm¹⁶.

Breakdown

How it tears

Most kneecap instability comes from a bending and twisting injury without contact, which pushes the kneecap outward¹⁷. Such an injury often also tears the soft tissues holding the kneecap from the inner side¹⁷. The 67-study review cites evidence that the MPFL ruptures once stretched beyond 12 to 18 mm¹. On that evidence, even a first dislocation would injure it¹.

A meta-analysis of 35 articles (2,558 patients) found MPFL injury in 94.7% of sudden outward dislocations³. It reported where the ligament tore³:

Share of injuries (%). Bars start at zero; the table below gives each value with its source.

Where it tore	Share of injuries
Kneecap end	37.1%³
Thigh-bone end	36.8%³
More than one site	25.1%³
Middle of the ligament	15.6%³

It tore most often at the kneecap end in people aged 16 or under, and at the thigh-bone end in those over 16³.

Complete and partial tears

Imaging and surgery studies report how badly it tears^18,19:

In 73 patients scanned within 7 weeks of a dislocation, 72 had an MPFL injury¹⁸. Of these, 51.4% were complete tears, most often at the thigh bone, and 48.6% partial¹⁸.
A systematic review of patients under 18 after a first dislocation found, in four studies of 248 knees, 47% complete tears and 41% partial¹⁹. No tear showed on MRI in 13%¹⁹.
A case series, a report with no comparison group, followed 74 children aged 9 to 14 after a first dislocation²⁰. Of these, 99% had an MPFL injury²⁰. Keyhole surgery found injury at the kneecap end alone in 81%, and a total rupture there in 66%²⁰.

How the shape of the knee plays a part

In the 73-patient study, tears at the kneecap end went with a larger TT-TG distance¹⁸. That distance measures how far outward the bony bump on the shin sits from the thigh-bone groove. The authors concluded that tear patterns depend on trochlear dysplasia, kneecap height and that distance¹⁸. Trochlear dysplasia is a shallow or badly shaped groove at the lower end of the thigh bone.

The groove’s shape also mattered in one cadaver study. In eight knees, a flat outer condyle (rounded knob) of the thigh bone cut the kneecap’s outward stability by 70% at 30 degrees²¹. The patellar dislocation page covers how the injury is defined, who it affects and what trials found.

Recovery

What happens to the torn ligament

The background section of one systematic review states that a damaged MPFL may heal if the knee is held still²². It adds that the ligament can end up longer and looser, leaving the kneecap unstable outward²².

The 67-study review cites one study where the MPFL was longer in unstable kneecaps (54 mm) than in normal ones (49 mm)¹. It was 64 mm in kneecaps that dislocated again and again¹. Other studies reported 47 to 72 mm in normal knees¹. The review notes it is not known which comes first¹. People may dislocate because the ligament was longer, or it may be longer because it healed that way¹.

A cohort study, which follows a group of people over time, took 53 patients managed without surgery after a first dislocation²³. Of these, 35 had torn it at the thigh-bone end²³. After 7 years on average, 13 of the 15 with an unstable kneecap had that type of tear²³. So did 8 of the 9 who had dislocated again²³.

When the ligament is rebuilt

MPFL reconstruction replaces the damaged ligament with a tendon graft (a piece of tissue moved to rebuild a damaged part)²². The graft is fixed in a tunnel drilled in the thigh bone and held with a screw²².

A systematic review of seven studies (380 knees) found that tunnel had widened in 38.7% to 77.1% of knees²². Evidence conflicted on how the width changes over time²². Five low-quality studies in that review found no worse knee scores with a wider tunnel²². The review noted the low level of that evidence²².

In ten cadaver knees, the intact MPFL held 208 newtons¹⁶. Stitched repairs and tendon grafts fixed in those knees held between 37 and 195 newtons, depending on the method¹⁶.

What ACL grafts show

The studies in this part are of grafts replacing the anterior cruciate ligament (ACL), not the MPFL. They do not show whether MPFL grafts change the same way. A systematic review of tissue-sample studies in people found an ACL tendon graft survives in the joint²⁴. It goes through ligamentisation, the slow change of a tendon graft into tissue like a ligament²⁴. The mature ACL graft may then look like a normal ACL under the microscope²⁴.

Mature ACL grafts still differ in how their finest strands of collagen, a strong rope-like protein, are laid out²⁴. There is no agreement on how long each stage takes²⁴. A review of 23 studies (706 patients) found ACL grafts still changing a year after surgery²⁵.

Connections

The joint and bones it joins

The MPFL steadies the kneecap within the knee joint, where the kneecap slides up and down in a groove on the femur^1,26. The MPFL holds the inner side, joining the femur to the patella¹.

A dissection found it also attaches under the lower end of the quadriceps mechanism⁷. In 12 cadaver knees, these upper inner restraints attached half on the quadriceps tendon and half on the kneecap⁹. A review notes this has led to the name “medial patellofemoral complex”²⁷.

The vastus medialis is the inner muscle of the quadriceps²⁶. Its lowest fibres are called the vastus medialis obliquus. In 50 knees, the MPFL merged with this part’s tendon before attaching to the kneecap⁴. In eight cadaver knees, the two sets of fibres meshed near the kneecap⁵. Relaxing this muscle in another eight knees cut the kneecap’s outward stability by 30%²¹.

Other bands share the job of holding the kneecap^2,21:

In nine cadaver knees at 20 degrees, the medial patellomeniscal ligament gave 13% of the restraining force².
In eight knees under quadriceps tension, tearing the inner retinacula cut outward stability most with the knee straight, by 49%²¹. The retinacula are bands of tough tissue at each side of the kneecap.

In the 50-knee dissection, the MPFL took fibres from the medial collateral ligament, the ligament on the inner side of the knee⁴. Related exercise pages: lunge, during which one study imaged its length in 15 healthy people¹⁵.