Furthermore, although our cohort is larger than others previously reported, the sample size of 36 patients still precluded us from performing multivariate adjusted analyses.