Present data are poor and exhibit limited internal and external validity; thus, direct comparison was not possible with the eligible studies. Therefore, a meta-analysis of present data may not serve as the basis for a general treatment recommendation but underlines the need of prospective, randomized, controlled clinical trials.