Validation of patellar height measurements on conventional radiographs and MRI in patients with patellofemoral instability
Author(s):
Tscholl P. (Switzerland)
Tscholl P. (Switzerland)
Affiliations:
ESSKA Academy. TSCHOLL P. 05/09/18; 209716; P12-1869 Topic: Anatomy
Dr. Philippe TSCHOLL
Dr. Philippe TSCHOLL
Login now to access Regular content available to all registered users.

You can access free regular educational content on the ESSKA Academy by registering as an 'ESSKA Academy User’ here

Access to Premium content is currently a membership benefit.

Click here to join ESSKA or renew your membership.
Abstract
Discussion Forum (0)
Rate & Comment (0)
Objectives: Patellar height is one of the most important risk factors for patellar instability. Several measurements are used in literature, the Caton-Deschamps (CD-I) and Insall-Salvati-Index (InSa-I) on conventional radiographs (cRx), and the patellotrochlear Index (PT-I) and the sagittal patellar engagement index (SPE-I) on MRI.
Comparison of the different measurement methods used on cRx and MRI interchangeably are lacking, as well as studies in a large cohort for patello-femoral indices on MRI.

Methods: 299 patients (344 knees) were collected from the in-house data base that underwent patellar stabilization surgery (either by MPFL-reconstruction or trochleoplasty) between 1998 and 2012.
One pre- and one postoperative imaging series was included of either cRx and MRI, since MPFL reconstruction and trochleoplasty are well known to decrease patellar height postoperatively.
620 cRx and 386 MRI were available for the CD-I and InSa-I measurements, 297 MRI were found where PT-I could be measured and 368 MRI for the SPE-I.
Youden's index was applied to establish optimal cut-off values of the InSa-I and CD-I measured on MRI. A CD-I >1.2 and an InSa-I >1.3 measured on cRx was considered a patella alta. Spearman test was performed to evaluate correlations. A p-value of <0.05 was considered to be statistically significant.

Results: The mean CD-I measured on cRx and MRI was 1.02±0.17 and 1.14±0.17, respectively. The mean difference was -0.10, p<0.01. The optimal cut-off value for the CD-I measured on MRI was 1.24, with a sensitivity of 68% and a specificity of 83%.
The mean InSa-I measured on cRx and MRI was 1.27±0.51 and 1.31±0.23, respectively. The mean difference was -0.06, p< 0.01. The optimal cut-off value for the InSa-I measured on MRI was 1.36, with a sensitivity of 71% and a specificity of 81%.
There was a significant difference in PT-I in the group with CD-I >1.2 vs. ≤1.2, 0.36±0.20 vs. 0.45±0.17, respectively (p< 0.05), contrary to the InSa-I group >1.3 vs. ≤ 1.3 (0.41±0.18 vs. 0.45±0.18, respectively (p=0.092)). The optimal cut-off value for the PT-I considering a CD-I of >1.2 as pathologic was 0.28 with a sensitivity of 85% and a specificity of 45%.
There was a significant difference in SPE-I in the groups with CD-I >1.2 vs. ≤1.2, 0.40±0.19 vs. 0.50±0.16, respectively (p< 0.01), and also in the groups with InSa-I >1.3 vs. InSa-I ≤1.3, 0.45±0.17 vs. 0.51±0.0.16, respectively (p< 0.01). The optimal cut-off value for the SPE-I considering a CD-I of >1.2 as pathologic was 0.44, with a sensitivity of 62% and a specificity of 60%, and for an InSa-I of >1.3 as pathologic was 0.40, with a sensitivity of 78% and a specificity of 41%.

Conclusions: CD-I and InSa-I measured on MRI slightly over-estimate the patellar height compared to the gold-standard measurements on cRx. The CD-I is preferably used over the InSa-I in the context of patellar instability since its cut-off value on cRx shows better conformity to the patellofemoral indices used on MRI.

Keywords:
Patellar instability, Patellar height, imaging, gold-standard, diagnostics
Objectives: Patellar height is one of the most important risk factors for patellar instability. Several measurements are used in literature, the Caton-Deschamps (CD-I) and Insall-Salvati-Index (InSa-I) on conventional radiographs (cRx), and the patellotrochlear Index (PT-I) and the sagittal patellar engagement index (SPE-I) on MRI.
Comparison of the different measurement methods used on cRx and MRI interchangeably are lacking, as well as studies in a large cohort for patello-femoral indices on MRI.

Methods: 299 patients (344 knees) were collected from the in-house data base that underwent patellar stabilization surgery (either by MPFL-reconstruction or trochleoplasty) between 1998 and 2012.
One pre- and one postoperative imaging series was included of either cRx and MRI, since MPFL reconstruction and trochleoplasty are well known to decrease patellar height postoperatively.
620 cRx and 386 MRI were available for the CD-I and InSa-I measurements, 297 MRI were found where PT-I could be measured and 368 MRI for the SPE-I.
Youden's index was applied to establish optimal cut-off values of the InSa-I and CD-I measured on MRI. A CD-I >1.2 and an InSa-I >1.3 measured on cRx was considered a patella alta. Spearman test was performed to evaluate correlations. A p-value of <0.05 was considered to be statistically significant.

Results: The mean CD-I measured on cRx and MRI was 1.02±0.17 and 1.14±0.17, respectively. The mean difference was -0.10, p<0.01. The optimal cut-off value for the CD-I measured on MRI was 1.24, with a sensitivity of 68% and a specificity of 83%.
The mean InSa-I measured on cRx and MRI was 1.27±0.51 and 1.31±0.23, respectively. The mean difference was -0.06, p< 0.01. The optimal cut-off value for the InSa-I measured on MRI was 1.36, with a sensitivity of 71% and a specificity of 81%.
There was a significant difference in PT-I in the group with CD-I >1.2 vs. ≤1.2, 0.36±0.20 vs. 0.45±0.17, respectively (p< 0.05), contrary to the InSa-I group >1.3 vs. ≤ 1.3 (0.41±0.18 vs. 0.45±0.18, respectively (p=0.092)). The optimal cut-off value for the PT-I considering a CD-I of >1.2 as pathologic was 0.28 with a sensitivity of 85% and a specificity of 45%.
There was a significant difference in SPE-I in the groups with CD-I >1.2 vs. ≤1.2, 0.40±0.19 vs. 0.50±0.16, respectively (p< 0.01), and also in the groups with InSa-I >1.3 vs. InSa-I ≤1.3, 0.45±0.17 vs. 0.51±0.0.16, respectively (p< 0.01). The optimal cut-off value for the SPE-I considering a CD-I of >1.2 as pathologic was 0.44, with a sensitivity of 62% and a specificity of 60%, and for an InSa-I of >1.3 as pathologic was 0.40, with a sensitivity of 78% and a specificity of 41%.

Conclusions: CD-I and InSa-I measured on MRI slightly over-estimate the patellar height compared to the gold-standard measurements on cRx. The CD-I is preferably used over the InSa-I in the context of patellar instability since its cut-off value on cRx shows better conformity to the patellofemoral indices used on MRI.

Keywords:
Patellar instability, Patellar height, imaging, gold-standard, diagnostics
Code of conduct/disclaimer available in General Terms & Conditions

By clicking “Accept Terms & all Cookies” or by continuing to browse, you agree to the storing of third-party cookies on your device to enhance your user experience and agree to the user terms and conditions of this learning management system (LMS).

Cookie Settings
Accept Terms & all Cookies