Cruciate ligament rupture and the role of radiological knee shape

Patients with an anterior cruciate ligament (ACL) rupture or a posterior cruciate ligament
(PCL) rupture, have a significant health burden; both for the patients themselves, and for
the general health system. With significant, short and long-term consequences. Loss of
mobility, loss of sports activity, increased risk of early osteoarthritis, and not returning
to their pre-injury sports capability are just some of the consequences. Therefore,
preventing cruciate ligament (CL) ruptures is both in the interest of patients, and in the
interest of the general health care systems. To prevent CL ruptures, risk factors need to
be identified. The overall aim of this thesis is to find risk factors related to the shape of
the knee, which can be used to select patients at risk for a CL rupture.
In chapter 2 we compared the lateral and Rosenberg view X-rays of 168 prospectively
followed patients with a ruptured ACL to a control group with intact ACLs, matched for
gender, after knee trauma. We used statistical shape modeling software to examine knee
shape and find differences in shape variants between both groups. In the Rosenberg view
X-rays, we found five shape variants to be significantly different between patients with
an ACL rupture and patients with an intact ACL but with knee trauma. Overall, patients
who had ruptured their ACL had smaller, flatter intercondylar notches, a lower lateral tibia
plateau, a lower medial spike of the eminence, and a smaller tibial eminence compared
to control patients. We concluded that in general, patients with an ACL rupture have
smaller intercondylar notches and smaller tibial eminences in comparison to patients
with an intact ACL after knee trauma.
In chapter 3 we retrospectively compared Magnetic resonance imaging (MRI) scans of 121
patients with a proven ACL rupture to 92 control patients with proven intact ACLs and
PCLs. Patients were selected for age, weight, height, and sex (by manual selection). We
measured the volumes of the intercondylar notch and ACL and PCL, the bicondylar width
(BW), the notch width (NW), and the notch width index (NWI). Second, we compared
the result between males and females. Patients with an ACL rupture had, on average, a
smaller NW (Psmaller volumes of the PCL (Paverage a smaller NW (P(P=.004) and PCL (Psexes (P=0.508). We concluded that a smaller notch dimension, smaller volumes of the
intercondylar notch, and smaller volumes of the PCLs are related to the presence of an
ACL rupture. Secondary, females have smaller volumes of the intercondylar notch, ACLs
and PCLs, but do not have a smaller NWI when compared to males.
112
Summary
In chapter 4 statistical shape modelling software was used to assess the shape of the knee
and determine any difference in anatomical landmarks. We compared the anteroposterior
(AP), lateral, and Rosenberg view radiographs of 94 patients with a ruptured PCL to a
control group of 168 patients matched by age, sex, and body mass index (BMI), but with
an intact PCL after a knee injury. We found shape variants on the AP and Rosenberg
view radiographs to be significantly different between patients who tore their PCL and
those with an intact PCL after a knee injury. Overall, patients who ruptured their PCL
have smaller intercondylar notches and smaller tibial eminences than control patients.
This chapter showed that differences in the shape of the knee are associated with the
presence of a PCL rupture after injury. A smaller and more sharply angled intercondylar
notch and a more flattened tibial eminence are related to PCL rupture. This suggests that
the morphology of the knee is a risk factor for sustaining a PCL rupture.
In chapter 5 our goal was to investigate if the morphology of the knee in patients who
ruptured their PCL may differ from that of control patients. The hypothesis was that the
intercondylar notch dimensions, 3-D volumes of the intercondylar notch, and the 3-D
volumes of both the ACL and the PCL were correlated to the presence of a PCL rupture.
The magnetic resonance imaging (MRI) scans of 30 patients with a proven PCL rupture
were compared to 30 matched control patients with proven intact ACL and PCL. Control
patients were selected from patients with knee trauma during sports but without cruciate
ligament injury. Patients have been matched for age, height, weight, BMI, and sex. The
volumes of the intercondylar notch and both the ACL and PCL were measured on 3D
reconstructions. Secondly, the bicondylar width, the notch width, and the notch width
index were measured for all subjects. The relationship between our measurements and
the presence of a PCL rupture was analyzed. The results show a significant difference in
the volumes of the intercondylar notch and the ACL between patients with a ruptured
PCL and control patients. Patients with a PCL rupture have smaller intercondylar notch
volumes and smaller ACL volumes. There were no significant differences in the bicondylar
width, notch width, and notch width index. In the control patients, a significant correlation
between the volume of the PCL and the volume of the ACL was found (0.673, pWe concluded that patients with a PCL rupture have smaller intercondylar volumes and
smaller ACL volumes when compared to control patients. Secondly, patients with smaller
ACL volumes have smaller PCL volumes. This study shows, for the first time, that there
are significant size and volume differences in the shape of the knee between patients
with a PCL rupture and control patients.
In chapter 6 we investigated whether the shape of the knee can predict the clinical
outcome of patients after an anterior cruciate ligament rupture. We used statistical shape
modelling to measure the shape of the knee joint of 182 prospectively followed patients
113
Summary
on lateral and Rosenberg view radiographs of the knee after a rupture of the anterior
cruciate ligament. Subsequently, we associated knee shape with the International Knee
Documentation Committee subjective score at two years follow-up. The mean age of
patients was 31 years (21 to 51), the majority were male (n = 121) and treated operatively
(n = 135). We found two modes (shape variations) that were significantly associated with
the subjective score at two years: one for the operatively treated group (p = 0.002) and
one for the non-operatively treated group (p = 0.003). Operatively treated patients who
had higher subjective scores had a smaller intercondylar notch and a smaller width of
the intercondylar eminence. Non-operatively treated patients who scored higher on
the subjective score had a more pyramidal intercondylar notch as opposed to one that
was more dome-shaped. We concluded that the shape of the femoral notch and the
intercondylar eminence is predictive of clinical outcome two years after a rupture of the
anterior cruciate ligament.
Finally, in chapter 7, the general discussion, some of the key findings of the studies in this
thesis are further discussed. Furthermore, we gave an overview of our studies, and how
they contributed to the current practice. The strengths and limitations were discussed.
To conclude, recommendations for practice and future research were presented.