Impact of Use of Intramedullary and Extramedullary Guides on Tibial Component Geometry in Total Knee Replacements: A Systematic Review and Meta-Analysis.

Mechanical guides in total knee arthroplasty are divided into intramedullary and extramedullary systems, designed to give accurate reference, to enable the surgeon to perform a tibial cut which is perpendicular to the mechanical axis. We conducted a systematic review and meta-analysis of levels 1 and 2 published data which directly compares the two methods of alignment, with outcomes of interest being the mean tibial component angle to the mechanical axis and the number of outliers from the optimal range. The PRISMA (preferred reporting items for systematic reviews and meta-analysis) guidance was followed. A search was conducted of online databases Medline PubMed; EMBASE; ISI Web of Science, and the Cochrane library, using the Boolean search string ([intramedullary OR extramedullary] AND knee AND [arthroplasty OR replacement]). Numerical data pertaining to tibial component alignment (TCA), the mechanical tibiofemoral angle, the tibial slope, and the number of outliers from optimal TCA were collated, and used to establish pooled results. No constraints on the search in terms of year of publication or language were instituted. Intrastudy bias was assessed using the Jadad score for randomized controlled trials and the Newcastle Ottawa score for prospective cohort studies. A total of 1,896 titles were reviewed. Following abstract review and full review of relevant articles, 10 publications were included for analysis, of which 8 were suitable to include for meta-analysis. No trials showed a significant difference in the mean TCA. Two trials showed an increased number of outliers in the extramedullary group and two studies showed an increased number of outliers in the intramedullary group. Pooled data from studies which included these outcomes showed no advantage for either system in limiting the number of outliers from the optimal TCA (relative risk, 0.99; 95% confidence interval [CI], 0.87-1.14; p = 0.004), and no significant difference in mean TCA (standardized mean difference, -0.07; 95% CI, -0.22 to 0.08; p = 0.000). Based on our results, no advantage can be attributed to the type of mechanical guide used in obtaining an adequate tibial cut.