This is a course on limb realignment, but limb alignment also affects the patella.

First published 2010, and reviewed August 2023 by Dr Sheila Strover (Clinical Editor)


The abnormal anatomy group is more intriguing for surgeons, and the first thing to do is to define the anatomical problems.

There are a number of specific anatomical abnormalities that we are on the lookout for -


Valgus alignment

Patients with abnormal anatomy are in valgus alignment, that is looking from the front, they really are quite knock-kneed. This means that the kneecap is already tending to slide off the lateral side of the knee.

Increased Q-angle

The patient often has what we call an ‘increased Q-angle’ (quadriceps angle). First you draw a straight line from the anterior superior iliac spine (the knobbly bit on the front of the pelvis) to the middle of the kneecap and down to the foot. And then you draw another line from the middle of the kneecap down the middle of your patellar tendon to your tibial tubercle, thereby creating an angle. In most people this angle is approximately 15 degrees. It is slightly greater in women than in men.

The Q-angle is one of those things that is very easy to talk about in clinic, it takes two seconds to do, but no-one actually draws lines and measures the angle with a goniometer.

Increased TTTD

These days we do have a very sophisticated way of measuring the Q angle and it is called the ‘tibial tubercle transfer distance’. Previously, we used CT scans to calculate this figure but that involved use of a lot of X-ray radiation. Now we can get the same information from an MRI scan (which does not involve any radiation). An MRI is made of digital slices and you can take the axial slices and use them to look at sections of the leg, like a tree that is being sliced from the top. The MRI can calculate very accurately how many millimetres it is from the middle of the trochlea to the middle of the tibial tubercle. Up to 18 mm is OK and anything over 20mm is really very abnorma. If the TTTD is over 20 mm and the patient is having serious problems with patella instability then surgery to reduce/normalise the TTTD is indicated.




Often when these patients sit with their leg dangling at 90 degrees of bend (flexion) and they straighten their leg up nice and slowly the kneecaps moves/tracks in an abnormal fashion. Instead of going up and down the leg (as you straighten and bend) it tracks towards the side and we call that J-tracking.


When we examine these patients they are often hypermobile, that is all of their joints are a bit too mobile (thumb, metacarpophalangeal joints of the fingers, they hyperextend at the elbow, they hyperextend at the knee).

Age and gender

Usually these patients have had problems from a very early age.  It is more common in females than in males, who may start to have trouble when they are 13 or 14 and occasionally even younger (the most common age is late teens or early twenties) They tend to complain of pain, and the pain comes from the instability.

Lateral tilt

When you look at the normal knee, the kneecap sits right over the front of the thigh. With someone with patellar instability usually their kneecap is actually tilting, pointing out to the outside.

Trochlear dysplasia

Once we have done our clinical assessment, we then do our radiological assessment and the critical X-ray view here is the lateral X-ray. From the lateral knee x-ray we can tell all sorts of things – we can tell the depth of the trochlear groove and we can tell its shape. In some individuals with a mis-shaped trochlea, developmental abnormalities can lead to either a flat or humped trochlea. The kneecap is therefore not sitting in a groove but sitting up on a hump. Both a flat and a humped trochlea fall under the umbrella of trochlear dysplasia. You can pick this up that on a lateral knee X-ray and this is really a critical investigation that can be carried out in in outpatients.

Patella alta

From the lateral X-ray again we then look critically at patellar height and we can see if the patella is low (patella baja or infera) or more commonly in these patients the patella is slightly high (patella alta).

So it is like building up on a points system, and when you get to a certain number of points you start running into trouble. So if  you are a teenage girl, hypermobile, in valgus alignment with an increased Q-angle, it won't take very much for the kneecap to suddenly slide off the side and dislocate.


Subluxator or dislocator?

These are two very different groups. There are patients who actually dislocate and then there are those ones who subluxate (where the knee cap drops out and then drops back in again). A very key part of the assessment is asking the question “Have you ever been to casualty (ER) with your kneecap stuck to the side of your leg and it’s had to be put back in again?” If the answer is “yes”, then the patient may require the surgery. If the answer is “no, it always pops back in by itself”, then the patient is unlikely to require surgery. Usually these patients settle with appropriate physiotherapy.


If a patient is a subluxator, we tend to avoid surgery and the key part of treatment is building up the VMO muscle. When examining a patient with an increased Q-angle, valgus alignment, and a laterally tilted patella, you can, just by placing your thumb on the side of the kneecap, push the kneecap laterally. The patients feel very apprehensive about this manoeuvre– and this is called the ‘apprehension sign’.

Now often the apprehension test is done with the leg fully extended which not the right way to do this test. It is amazing how many consultants do this test with the leg in extension, and have no idea that this an incorrect technique. It makes absolutely no sense to examine the kneecap with the leg fully extended as the kneecap is nowhere near the trochlear groove in this position - it is sitting two or three centimetres north. If you sit with your leg completely relaxed you can move your kneecap anywhere you like. It is when you put the knee at 30 degrees that the kneecap engages in the groove - this is the position in which you should examine the patient. You need to examine the leg at 30 degrees and see how much apprehension and how much lateral displacement there is. In a normal knee we divide the width of the kneecap into 4 quadrants. So halfway is two quadrants. If you draw a line right the way down the middle of the kneecap it should go medially up to two quadrants and it should go laterally up to two quadrants. Now if you have ruptured your MPFL or you have got very abnormal anatomy, having also stretched all the medial structures because of the previous subluxations, the kneecap can be moved to the outer/ lateral side by more than two quadrants and you can sometimes just take the kneecap off to the side. The reverse may be true – they may be very very tight laterally so that when you try and displace the patella medially it doesn’t even go a quadrant – that is a very tight patella and that is probably the only indication there is for a lateral release (which should be an extremely rarely done operation).

So if you examine the patients and they’ve got some pain, their kneecap has virtually no movement medially and perhaps some excessive movement laterally  - if everything else is intact and the patient is not actually dislocating then there is a very small percentage of people you should recommend the lateral release to. A very busy knee surgeon who is seeing hundreds of patients every year and doing hundreds of operations, may only do one lateral release per year. It is therefore an extremely rare scenario.

So we have talked a little about the patient presenting with pain and instability, and about the fact that the instability is either the kneecap just sort of going out and in by itself (subluxing) or staying out (dislocating). It is very important to get that clear because people often think their kneecaps are dislocating when they are not. They are just subluxating.


If there is clear trochlear dysplasia, ie. the patient has a hump instead of a V on the lateral X-ray, then a failure of conservative management should lead to an operation called trochleoplasty. The knee is opened up and the end of the trochlea is refashioned to create a new groove for the patella to sit in. This operation can only be done once you have reached skeletal maturity, and the cut-off age where that operation tends to stop being done is about 40. Once these patients get beyond their mid-30s they tend to have too much wera and tear in the patello-femoral joint to really benefit from that procedure. Trochleoplasty has not been around for a very long and it is only really done in three centres in the UK at the moment - Norfolk&Norwich, Warwick&Coventry and Bristol. This is where we all refer our patients to.

The trochleoplasty that used to be done is what was called an LV procedure. That was the old-fashioned trochleoplasty where you basically just took an osteotome to the side of the femur and wedged it up – so that was a form of trochleoplasty. There is now a  new procedure that has been around  for only the last five to ten years. It was developed mainly by the French surgeons in Lyon  (Henri Dejour and now his son David Dejour) and they probably have the biggest surgical experience in the world with patellofemoral instability. Their group which includes Philippe Neyret and Michele Bonin are probably the key players in Europe, and in the UK now Jonathan Eldridge and Simon Donnell head the group. Pete Thompson in Coventry and Warwickshire has made his name in patellofemoral instability, especially in MPFL reconstruction. Between the three of them we probably have the top experts in trochleoplasty, with no-one else really tackling that operation. Most of the knee surgeons in the UK would refer our trochleoplasties to these three units. Most knee surgeons are happy with the MPFL reconstruction and tibial tubercle transfer but trochleoplasty for dysplasia (which is a very rare problem) needs really to be done in those centres. That is what most of us believe.


Lateralised Tibial Tubercle

The other more common abnormal anatomy is when the trochlea is OK but the tibial tubercle is offset too far laterally. This can be addressed with the so-called Fulkerson osteotomy or modification of that particular procedure. Whatever you want to call it, and there are a number of different names, the idea is that you osteotomise the distal insertion of the patellar tendon (at the tibial tubercle) with a 4 cm block of bone and you transfer and fix it medially. The transfer distance is calculated on the MRI scan that you do pre-operatively. This operation has been given lots of different names and there have been lots of different variations on the original procedure. elmslie-trillat procedureThe most commonly done operation now is a so-called modified Elmslie-Trillat or Fulkerson procedure. What that procedure does is it not only moves the tibial tubercle medially but the osteotomy is carried out in such a way that you move the bone slightly anteriorly so that not only are you pushing it medially but you are pushing it anteriorly. If you imagine if you cut the tibia from the lateral side if you cut it completely flat and you push the bone, the bone will just move in a flat plane. If you drop your hand and you cut it up a hill, you slide that piece of bone uphill. Conversely if you lift your hand up and you cut downhill, you can slide that piece of bone downhill. What we know is by dropping our hands and pushing it upwards we take pressure off the patellofemoral joint by moving everything anteriorly. If you do it flat it is ok, if you actually go downhill it is bad news because you are actually bringing too much pressure into the patellofemoral joint so what we tend to do is drop our hand from the lateral side, cut the bone in such a way that we slide the piece of bone uphill and then we fix it with a couple of screws. The problem with that procedure is that if you ask most surgeons “How far do you push that piece of bone?” you tend to get the same response and that is “A centimetre”.

Now obviously everyone is going to be slightly different. When I was working in Brisbane we researched a very novel idea that one of the surgeons, Peter Myers had and which he worked on with Andy Williams. Their idea was to stimulate the femoral nerve during the operation, which would make quadriceps muscle contract and pull on the patella, giving you some sort of feel for when the patella is in the middle of the trochlear groove.

Initially when you stimulate the femoral nerve the patella will jump off to the side laterally because it is tilted and it is sitting laterally. When you do your osteotomy and you get the patella to sit in the middle and then you stimulate the femoral nerve, the patella will not go too far laterally or go too far medially – it will tend to sit in the middle. And that was a very nice idea, but this is something that is quite difficult to achieve during the operation. But that is now how I do it – we do intraoperative femoral nerve stimulation and that gives us a slightly more scientific feel about how far we should be moving the tibial tubercle. It might be slightly imprecise but it works extremely well. This tends to be a bilateral problem and the patient will often ask: “When can you do my other side?”, so it is a very successful operation.