Comparison of transcatheter, minimally invasive and conventional surgical aortic valve replacement: a systematic review and network meta-analysis

Type of funding sources: None.

The landscape of aortic valve replacement (AVR) has evolved dramatically over the years, but long-term outcomes have yet to be comprehensively explored.

To compare long-term mortality among three AVR techniques: transcatheter (TAVI), minimally invasive (MIAVR), and conventional surgical AVR (CAVR).

An electronic literature search was performed for randomized controlled trials (RCTs) comparing TAVI to CAVR, and RCTs or propensity score-matched (PSM) studies comparing MIAVR to CAVR or MIAVR to TAVI. Individual patient data for all-cause mortality was derived from graphical reconstruction of digitized Kaplan-Meier curves. Pairwise comparisons followed by network meta-analysis were conducted. Sensitivity analyses were performed in the TAVI arm for high risk and low/intermediate risk as well as transfemoral (TF-TAVI) patients.

A total of 27 studies involving 16,554 patients were included. In the pairwise comparison of TAVI versus CAVR, TAVI showed superior mortality to CAVR until 39.5 months, beyond which there was no significant difference in longer term mortality. When restricted to TF-TAVI versus CAVR, consistent mortality benefit favoring TAVI was seen (shared-frailty hazard ratio [HR]=1.17, 95%CI=1.03-1.33, p=0.016). In the network meta-analysis involving majority PSM data, MIAVR was associated with significantly lower mortality than TAVI (HR=0.69, 95%CI=0.59-0.82) and CAVR (HR=0.68, 95%CI=0.58-0.80); this significant association was not seen when compared to TF-TAVI patients.

An initial short-medium term mortality benefit for TAVI over CAVR was noted but this benefit was attenuated over the longer term. In the subset of TF-TAVI patients, this long-term mortality benefit persisted, suggesting that non-TF techniques are associated with higher mortality. Amongst majority PSM data, MIAVR showed improved mortality compared to TAVI and CAVR but this was not seen in the TF-TAVI subset. This suggests that MIAVR may have a role in treating patients who are ineligible for TF-TAVI in experienced centers, and future RCTs are needed to conclusively validate this.