Clinical anatomy of the anterior cruciate ligament and pre-operative prediction of ligament length

Background: Ligament grafts used in anterior cruciate ligament (ACL) reconstruction need to be the correct length for proper functioning. If the graft length is incorrect, the patient could risk knee instability, loss of range of motion, or failure of graft fixation. Easier and time-efficient reconstruction will be facilitated if the length of the ACL is predicted in advance. Apart from examining the morphological properties of the ACL, this study aimed to determine whether the epicondylar width of an individual can be used to predict ACL length and thereby assist in restoring the normal anatomy of the ACL. Methods: Ninety-one adult cadavers were studied. Patellar ligament (PL) length, ACL length, ACL width and the maximum femoral epicondylar width (FECW) were measured. Results: The morphology of the ACL and PL was determined. The results revealed that FECW was the most reliable predictor of ACL length. A linear regression formula was developed in order to determine ACL length by measuring maximum FECW. Conclusions: ACL and PL morphology compared well with the results found in previous studies. An individual’s FECW can be used to predict ACL length pre-operatively. These results could improve pre-operative planning of ACL reconstruction.


Clinical anatomy of the anterior cruciate ligament and pre-operative prediction of ligament length
Mrs R van Zyl, 1,2 MSc Anatomy Dr A-N van Schoor, 2 PhD Anatomy Dr PJ du Toit, 3 PhD Physiology Dr EM Louw, 4 PhD Statistics

Results
All data were symmetrically distributed.The descriptive analysis results are shown in Table I.The Pearson's correlation coefficient results can be seen in Table II.The Pearson's correlation coefficient r-value was determined to suggest which of the four independent variables demonstrated a significant positive or negative linear relationship with the dependent variable, ACL length.
(Table II).If the p-value of the Pearson's correlation coefficient test was p<0.05, it was considered as a statistically significant correlation.Statistical inference followed after the completion of the descriptive statistics in order to predict ACL length.By means of backward elimination using a multiple regression model, it was observed that both height and FECW had a statistically significant correlation to ACL length.However, FECW was a more reliable predictor for estimating ACL length than height (r=0.36 against r=0.31,respectively).On the basis of this estimation, no additional variables were used to compile an equation for preoperative ligament length estimation.A linear regression formula was developed for predicting ACL length and, therefore, optimum final graft length to re-establish the intra-articular length of the torn ACL.The following equation (Equation 1) can be used to predict ACL length by measuring the maximum FECW: ACL length = 0.25 (FECW) + 11. 18  (1) where ACL=anterior cruciate ligament; FECW=femoral epicondylar width

Discussion
The mismatch of grafts can be minimised by more accurate prediction of ACL intra-articular length and the required graft length for ACL reconstruction.The first aim of this study was to determine the morphology of the ACL and PL.This study has shown that the mean intra-articular ACL length of this sample was 32.4 mm, with a range between 22.9 and 45.1 mm.The literature has shown different results for the mean length of the ACL, namely: 20 mm, 24 31 mm, 25 32 mm, 5,7,9 38 mm, 8 and 39 mm. 26The ACL length has exhibited some variability, with reported ranges of 20-25 mm, 13 22-41 mm, 9 22-44 mm, 7 25-35 mm, 25 and 37-41 mm. 26The mean width of the ACL calculated in the current study was 8.2 mm, ranging from 5.1-13.7 mm.This measurement differs slightly from the range of 7-12 mm as determined by Bicer et al., 7 Zantop et al. 5 and Odensten and Gillquist. 25Previous authors found the ACL to have a mean width of 10 mm 25 and 11 mm. 8he PL mean length of this sample was 45.7 mm, ranging between 31.6 mm and 59.2 mm.These measurements were similar to results from previous studies: Brown et al. 13 recorded a range of 45-50 mm and Denti et al. 24 a mean length of 45.5 mm.On average, the PL is 14 mm longer than the ACL in this study and is therefore sufficient in length to allow effective fixation of the graft.
Variation in PL length could explain the occurrence of graft mismatch found with ACL reconstructive surgery.Lastly, the slight differences between both the ACL and PL measurements from this sample and other reported measurements could be attributed to the respective population differences of the samples studies.The comparisons were made between the current sample (South African), an Italian sample 24 and a North American sample. 13The measuring technique used could also account for the dissimilarities found among the studies referred to by Brown et al. 13 and Denti et al. 24 ACL length and PL length demonstrated no significant statistical correlation (p=0.92), as was the case between ACL length and ACL width (p=0.26).The significant correlation between ACL length and height (p=0.01) was not unexpected.Brown et al. 13 in 2007 found a strong correlation on a large magnetic resonance imaging (MRI) study sample consisting of 414 knees, and concluded that patient height can predict the required length of the BTB graft used for ACL reconstruction.In contrast, Denti and coworkers, 24 using 50 reconstructed and nine cadaver knees, were unable to confirm this relationship between ACL length and patient height.ACL length and FECW revealed a significant correlation of p<0.01 in this study.It was also statistically determined that FECW could account for the other three independent variables in predicting ACL length as FECW had a strong correlation with all other variables.Femoral epicondylar width was found to be the most reliable predictor of ACL length (r=0.36) as opposed to height, which had an r-value=0.31.A significant model was developed for the estimation of ACL length when FECW is known.The R 2 -value of 0.13 indicated that 13% of variation found in ACL length (among the cadavers) could be accounted for by Equation 1.Therefore, measuring the FECW of the patient might allow for pre-operative estimation of the length of the ACL by using the previously mentioned linear regression formula.These values could then be utilised to predict the required graft length to replace the torn or ruptured ACL.

Variables
Maximum FECW was used in the current study to correlate with ACL length.Patient height is often unknown and FECW is a practical measurement, as it is commonly available on X-ray and MRI. 13,24In this sample, FECW has also proven to be a slightly more reliable predictor than height for ACL length estimation.When comparing ACL length and FECW, a definite positive correlation between the measurements (p<0.05) was noticed.However, the low R 2 -value of the linear regression indicates that in the South African cadaver sample examined in the current study, the FECW cannot be the sole predictor of ACL length.Preoperative mismatch of the graft length could occur if the FECW is measured without the incorporation of additional imaging modalities.Intra-operative measurements may also assist to determine ACL length and therefore the required graft length.
The weak positive correlation observed for predicting ACL length using the maximum FECW could be explained by remnants of soft tissue covering the epicondyles after dissection, which may have led to inaccurate measurement of the maximum FECW.Dissection of the ACL was also problematic because the ligament coursed diagonally within the knee joint.The measurement was therefore taken blindly, which could have resulted in inaccurate measurements of the true ACL length.Utilising imaging modalities could allow clinicians a greater degree of accuracy in estimating ACL length from measuring the FECW of the patient.Future studies could therefore be designed to investigate whether the correlation between the ACL length and the FECW could be improved when measured on MRIs.These findings should be verified in the clinical setting.

Conclusion
The anatomy of the ACL and PL of this South African cadaver sample compared favourably with results reported in previous studies conducted on other population groups.Similar to some of these studies, a correlation was found between the intra-articular ACL length and the height of the cadaver.It was also found that the maximum FECW of a patient can predict the length of the intra-articular ACL more reliably than the height of the patient can.Using FECW as a possible predictor provides an additional method for estimating the pre-operative length of the graft required for ACL reconstruction and minimising the chance of graft mismatch.

table ii : Correlation matrix to indicate associations between the dependent (ACL length) and independent variables
a Statistically significant correlation ACL=anterior cruciate ligament; FECW= femoral epicondylar width; PL=patellar ligament