Read on Twitter
Just published in #JACC
Multi-centre propensity score-matched observational study from the Redo-TAVR registry (initiated in Feb '19) comparing TAVI-in-TAVI versus TAVI-in-SAVR!
A thread exploring this paper...
Multi-centre propensity score-matched observational study from the Redo-TAVR registry (initiated in Feb '19) comparing TAVI-in-TAVI versus TAVI-in-SAVR!
A thread exploring this paper...
37 centres performed 63,876 TAVI procedures from April '05 - April '19 of whom 434 (0.7%) had already previously undergone TAVI. Importantly, 223/434 (51%) were done "urgently" as bailout procedures and were *excluded* from the analysis.
That's over 1/2 the group excluded...
That's over 1/2 the group excluded...
After propensity score matching, n=165 in each group.
Mean age ~80. 60% males. Mean STS score 6.5-6.8%. ~62% NYHA Class III, ~16% NYHA Class IV.
Repeat intervention for pure AS, pure AR or mixed AS / AR was roughly 1/3 in each group.
Self-expandable TAVI valve in ~60% cases
Mean age ~80. 60% males. Mean STS score 6.5-6.8%. ~62% NYHA Class III, ~16% NYHA Class IV.
Repeat intervention for pure AS, pure AR or mixed AS / AR was roughly 1/3 in each group.
Self-expandable TAVI valve in ~60% cases
RESULTS
- There was a technical efficacy endpoint & a safety endpoint, both of which, as usual, were multi-composite outcomes.
- Procedural / technical success achieved in 72.7% TAVI-in-TAVI versus 62.4% TAVI-in-SAVR, p 0.045, driven mostly by residual mean gradient >20mmHg
- There was a technical efficacy endpoint & a safety endpoint, both of which, as usual, were multi-composite outcomes.
- Procedural / technical success achieved in 72.7% TAVI-in-TAVI versus 62.4% TAVI-in-SAVR, p 0.045, driven mostly by residual mean gradient >20mmHg
- No significant difference in mortality, stroke, MI, coronary obstruction, new pacemaker or length of stay.
- Major bleeding nearly double in TAVI-in-TAVI group (10.3% vs. 5.2%) with borderline p value (0.061)
- More residual AR at 30 days & at 1yr with TAVI-in-TAVI
- Major bleeding nearly double in TAVI-in-TAVI group (10.3% vs. 5.2%) with borderline p value (0.061)
- More residual AR at 30 days & at 1yr with TAVI-in-TAVI
DISCUSSION
This is a very useful hypothesis-generating study. As the number of TAVI procedures
, especially with ACC/AHA giving TAVI a Class I recommendation for age >65, it's inevitable that we WILL see more patients with TAVI degeneration requiring re-do procedures
This is a very useful hypothesis-generating study. As the number of TAVI procedures
, especially with ACC/AHA giving TAVI a Class I recommendation for age >65, it's inevitable that we WILL see more patients with TAVI degeneration requiring re-do procedures
We know from a different recent study that open SAVR after a prior TAVI carries high risk (mortality >20%), so understanding TAVI-in-TAVI outcomes will be crucial
All usual limitations apply - registry study, no core lab for imaging, unmeasured biases not corrected by propensity score matching etc.
Also, excluding HALF the group that had TAVI-in-TAVI just because it was urgent is problematic...5 or 10% ok maybe, but excluding 50%! Why?
Also, excluding HALF the group that had TAVI-in-TAVI just because it was urgent is problematic...5 or 10% ok maybe, but excluding 50%! Why?
I can only imagine it's because their outcomes were worse so data would look less impressive? Have I misunderstood this n=223?
I see this study as showing that in selected patients, TAVI-in-TAVI is safe & has reasonable 1yr outcomes. Clearly longer term outcomes needed...
I see this study as showing that in selected patients, TAVI-in-TAVI is safe & has reasonable 1yr outcomes. Clearly longer term outcomes needed...
Finally, something no-one talks about-cost! How many countries could afford to treat the average 65 or 70yr old with 2 TAVIs? I found out how much we pay for valves, it's eye-opening:
Mechanical AVR - £1100
Perimount Magna - £1500
Edwards Inspiris - £3000
Sapien 3 TAVI - £21000
Mechanical AVR - £1100
Perimount Magna - £1500
Edwards Inspiris - £3000
Sapien 3 TAVI - £21000
Of course the large difference in cost of valve is compensated for by avoidance of ICU, less blood transfusions, less dialysis & shorter hospital stay...does that fully offset the cost differences? I don't know the answer to that...
Conclusion that TAVI-in-TAVI had better procedural success may be statistically true - just - but is a bit enthusiastic given actual data. Eg the mean gradient difference is 12.6mmHg vs. 14.9mmHg, p=0.01. Probably a good example of statistical but not clinical significance, imho
Congrats to authors though, very useful and novel data that will need longer term follow-up and detailed scrutiny, eg does the type of TAVI valve you use in 1st & 2nd procedure (for TAVI-in-TAVI) affect outcome and, if yes, how?
#cardiotwitter #MedTwitter
#cardiotwitter #MedTwitter
A clarification after discussion with @djc795 - looks like excluded n=223 were patients who had TAVI that had an intra-procedural problem that then had *immediate* 2nd TAVI procedure (i.e. at same sitting as 1st procedure). In this case, agree with their exclusion in *this* study
Reason this didn't occur to me immediately is that, fortunately, I've never seen a 2nd TAVI deployed inside a TAVI valve all in one procedure! So it didn't occur to me this could have happened 223 times! But, as pointed out, this data goes back to 2005 (1st TAVI was only in 2002)
More thoughts! One problem with 15yr timeframe in a TAVI study is that TAVI technology has improved so much & continues to evolve. Thus, although major bleeding almost double for TAVI-in-TAVI, is that mostly from early times with much bigger femoral sheath sizes?
