Risk factors for leg length discrepancy in patients with congenital vascular malformation

Objectives

This study was conducted to determine the clinical risk factors for the development of leg length discrepancy (LLD) in patients with congenital vascular malformation (CVM) affecting the lower extremity.

Methods

A retrospective analysis was conducted of a prospectively collected database that included 361 patients who underwent assessment of a CVM lesion from September 1994 to January 2005. We measured LLD using lower extremity scanograms of 229 patients who were suspected of having LLD on physical examination. The risk factor analysis for a clinically significant LLD (>2 cm) was performed with the variables of age, gender, features of CVM (type, extent and depth), and the deep vein status (agenesis, hypoplasia, phlebectasia) of the affected limb. Fisher’s exact test and the χ² test were used for the univariate analysis, and a logistic regression test was performed for the multivariate analysis. Among the patients with LLD, we compared the overgrowth group and undergrowth group with the Fisher’s exact test to identify differences between the two subgroups.

Results

The included patients were 153 males (42%) and 208 females (57%) with a mean age of 20 ± 14 years (range, 1 to 62). There were 157 patients (43%) in the still growing age group (age <15 years) and 204 (57%) were in the finished growing age group (age >15 years). The types of CVMs included 215 venous (60%), 43 arteriovenous (12%), 46 lymphatic (13%), and 57 (16%) combined venolymphatic malformations. On the lower extremity scanogram, 26 patients (7%) had a LLD of >2 cm due to overgrowth (n = 20) or undergrowth (n = 6) of the affected limbs. The univariate analysis showed that a venolymphatic malformation (P = .003) and a whole leg CVM (P = .000) were significant risk factors for development of LLD. However, the multivariate analysis identified the whole leg CVM lesion as a single independent risk factor for LLD (P = .004; odds ratio, 6.512, 95% confidence interval, 1.788 to 23.713). On subgroup analysis, a whole leg CVM was also identified as a risk factor for overgrowth of the affected limb. In a comparison between two subgroups of LLD (overgrowth v. undergrowth), overgrowth was significantly (P = .022) more common in female than in male patients.

Conclusions

As a clinical risk factor for development of LLD, the extent of the CVM lesion was a single independent risk factor regardless of the type or depth of the CVM lesion. In addition, our data suggest that overgrowth or undergrowth of the affected limb, as a cause of LLD, might be related to gender.