The unbeatable 5 - Warzee et. al.

Effect of Tibial Plateau Leveling on Cranial and Caudal Tibial Thrusts in Canine Cranial Cruciate-Deficient Stifles: An In Vitro Experimental Study.

Christine C, Warzee, Loic M. Dejardin, Steven P. Arnoczky, Ruby L. Perry

Veterinary Surgery 30:278-286, 2001

The researchers used fifteen pelvic limbs from adult dogs of similar size and body weight (range, 27 to 36 kg), which were euthanised for reasons unrelated to the study.

The limbs were mounted on a device with the stifles and hocks kept at 135 degrees and 145 degrees of flexion, respectively. This was achieved by using a tie between the femur and tibia, and between the calcaneus and the femur, mimicking the quadriceps and gastrocnemius muscles, respectively. The TPLO was performed, and the proximal tibial fragment could be rotated through predetermined degrees using a custom-made plate that functioned like a hinge. A thrust was induced in the stifle joint by compressing the femur towards the tibia.

Cranial or caudal subluxation of the tibia relative to the femur was measured using radiographic markers impended in the bones. Tibial axial rotation and strain in the caudal cruciate ligament were also measured. To measure the strain in the caudal cruciate ligament, the researchers used five specimens in addition to the fifteen limbs they used for all other measurements.

The limbs were placed under compression, and the cranial cruciate ligament was transected. The researchers found that this induced a mean cranial tibial subluxation of 18.9 mm. The tibial plateau was then rotated until the tibia was reduced. The mean Tibial Plateau Angle (TPA) that resulted in the reduction of the tibia, was 6.5 degrees. At this point in time, a caudal tibial subluxation with a mean value of 3.2 mm was also recorded. In other words, rotating the plateau to a TPA of 6.5 degrees reduced the tibia but also produced a caudal subluxation of 3.2 mm. This subluxation increased to a mean of 6.3 mm when the TPA was further decreased to 0 degrees. With the TPA at 0 degrees, the caudal cruciate ligament was transected, resulting in an even greater caudal translation of 8.9 mm.

From the above, it was concluded that to reduce the tibial subluxation of cranial cruciate ligament-deficient stifles, the tibial plateau needed to be rotated to a mean angle of 6.5 degrees that produced a caudal tibial thrust and that further rotation of the plateau resulted in an increase of this caudal thrust.

The researchers also found that transection of the cranial cruciate ligament resulted in internal tibial rotation, which decreased but was not eliminated after a TPLO to a TPA of 0 degrees. The researchers do not specify whether these measurements were performed while the construct was under a compressive load inducing a thrust. So, we assume this was the case.

Lastly, it was found that in the five specimens used to measure strain in the caudal cruciate ligament, strain increased by 37.7% when the TPA was decreased from the minimum required to eliminate cranial tibial translation to 0 degrees.

In my opinion, this is a very well-designed, performed and written study. Nevertheless, I would like to dig into the numbers more and one of the conclusions. I will start with the latter.

The second objective of the study was to find the minimum TPA that would eliminate the cranial tibial thrust. However, by design, this is not what it was measured.

Let's name the force that causes cranial tibial subluxation, cranial tibial thrust, and the one that causes caudal subluxation, caudal tibial thrust. When the limbs were loaded, the compression within the joint induced a cranial tibial thrust, resulting in cranial tibial subluxation in the cranial cruciate ligament-deficient stifles. A caudal thrust that is greater than the cranial thrust needs to be applied to reduce the tibia. If the two thrusts are equal in magnitude, the bone will not move. So, the only way that the tibiae of this experiment could be reduced was by rotating the plateau to an angle that would induce a caudal thrust greater than the cranial thrust. In other words, the plateau had to be mildly over-rotated compared with the rotation that would be required to only eliminate the cranial tibial thrust, which is what we try to achieve clinically. Clinically, we want to rotate the plateau until the caudal thrust equals the cranial thrust. Conversely, in this study, the plateau had to be rotated so the caudal thrust was mildly higher than the cranial thrust. This resulted in mild caudal subluxation of the tibia and strain in the caudal cruciate ligament, even when the plateau was rotated to its minimum required angle for reducing the tibia.

One would say that this does not matter because even if the caudal cruciate ligament is not loaded when the stifle is at 135 degrees of flexion, it may start being loaded in different angles of flexion, which means that the TPLO will cause an overall increase in the strain of this ligament. So, although the researchers had to induce a mild caudal tibial thrust to begin with, the caudal thrust would be there anyway during the stance phase of gait, even if it is not there at an angle of 135 degrees.

Although this may be true, it remains to be studied. Also, the above means that the minimum TPA may be higher than 6.5 degrees, which is something that was not discussed in the study.

We all know the challenges of fitting everything regarding your study in a manuscript with strict word count restrictions. So, the above is not a critique against the study but an addition to the limitations of the study.

The most important aspect of this study I would like to look into is some of the numbers without getting into too much statistical complexity. The authors present the mean TPA angle that achieved a reduction of the tibia as a target value. We are all used to normal reference ranges for the various measurements in blood samples, and we know the reason for a reference range is that we cannot expect all animals to have the same normal value. So, a question is how useful is the mean angle as a target clinically? This brings up another question. What was the range of the minimum TPA of this study? Unfortunately, the authors do not provide the range. To give an example of why this may be important, I will come up with a sample of 15 measurements with a mean of approximately 6.5 degrees and a standard deviation of 0.9 degrees as the sample in the study. The values of such a sample are:

5, 5.5, 5.9, 6.2, 6.3, 6.5, 6.7, 6.8, 7.1, 7.0, 7.0, 7.0, 7.0, 9

We notice that five values of our imaginary sample are below the value 6.5 degrees. But we know that this is just another possible sample of a population of dogs of the same size and body weights between 27 and 36 kg. We happened to choose this sample by chance. What if we had chosen a different sample? What would be the values then? Can we calculate a possible range of minimum TPAs a random dog from the above population would have?

Yes, we can. Such a range is called the prediction interval, and if we calculate it by setting the confidence level at 95%, it means that there is approximately 95% probability that a random dog weighing between 27 to 36 kg would require a minimum TPA within this interval. We can do this based on the sample size, mean and standard deviation of the study's sample, assuming the dogs in the sample were not chosen with a bias regarding the tendency of cranial tibial thrust.

I have calculated the prediction interval, and it is 4.4 to 8.4 degrees. So, a randomly chosen dog weighing between 27 and 36 kg of the size similar to the size of the dogs in the study has a 95% probability that it will require a TPA between 4.4 to 8.4 degrees for the cranial tibial thrust to be mitigated. There is a 2.5% probability that the minimum TPA is lower and a 2.5% probability that it is higher than the interval. If we would like to err to the side of caution, we could say that the minimum TPA required to mitigate cranial tibial thrust in 97.5% of a population of such dogs is 4.4 degrees.

So, have we defeated the unbeatable 5 degrees rule? Not yet! We said that the researchers found that a minimum TPA angle of 6.5 degrees was required to convert a cranial tibial translation to a caudal one. We also said that this would be an over rotation in comparison to what we would clinically require. So, let's say that the minimum TPA required is approximately 7 degrees. Using this as a mean value in a sample of size 15 with a standard deviation of 0.9 degrees, we can calculate a new prediction interval. This time, the interval is 5 to 9 degrees!

So, now we can say that the minimum TPA required to mitigate cranial tibial thrust in 97.5% of a population of such dogs is 5 degrees.

I don't know if this is coincidence or karma.

Using the data from the study that proposes 6.5 degrees as the target TPA after a TPLO, we calculated with 95% confidence that a random dog would need a minimum of 4.4 degrees, but considering the 6.5 found in the study is a mild over-rotation, we ended up with a minimum TPA of 5.

The unbeatable 5!