Thread by @dr_benoy_n_shah

#ESCCongress

Let's get straight to the point about #INVICTUS

The only people not surprised by the results are imagers that do TOEs (TEEs) on these patients! No way VKA was going to lose to NOACs here!😂

A 🧵...

#medtwitter #cardiotwitter #echofirst #epeeps

#ESCCongress
Patients with rheumatic MV disease typically have very different atrial appearances on TOE. No matter how well anticoagulated, there is very often spontaneous echo contrast in the LA

These atria are just different...why?

Who is this?

The answer is Ludwig Aschoff (1866 - 1942), German physician & pathologist

He described what are now known as Aschoff bodies - inflammatory infiltrates in the atrial walls that eventually turn to fibrotic tissue, which contributes to the atrial myopathy we see in these patients

#ESCCongress
Anyway, back to the trial!

Now, we know NOACs have completely changed the way non-valvular AF is treated in many countries. VKAs rarely given now in UK for newly diagnosed AF, most get a NOAC

But 'valvular AF' omitted from landmark RCTs

#ESCCongress
Now, back in 2019, @ToddNeale reported on behalf of @TCTMD that a retrospective observational S Korean study suggested a benefit for NOACs in mitral stenosis patients

However, discussants wisely suggested this needed to be tested in a RCT

tctmd.com/news/doacs-pat…

#ESCCongress
So, here comes the INVICTUS trial

Investigation of Rheumatic AF Treatment Using Vitamin K Antagonists, Rivaroxaban or Aspirin Studies

Question - is it safe to give patients with rheumatic MV disease & a history of AF a NOAC instead of a VKA?

Non-inferiority trial

#ESCCongress
INCL criteria:
Age >18
Echo evidence of RHD
AF/flutter at any time in the past

PLUS ONE OF THE FOLLOWING:
CHADS-VASc score 2 or more
MV area <2.0cm2 on echo
LA spontaneous echo contrast / thrombus on echocardiography

85% had mod-severe MS. ~1/4 had MVA <1.0cm2

#ESCCongress
138 recruiting sites across 24 countries
43% Africa, 40% Asia, 8% Latin America, 9% others
4565 patients recruited

Mean age 50
72% female

#ESCCongress
Trial recruited from Aug 2016 - Sept 2019

Original primary EP had to be changed during the trial due to fewer than expected strokes but more than expected deaths

Primary EP - composite of stroke/systemic embolism/MI/death

Details are here...

Patients randomised to rivaroxaban got 20mg or 15mg OD based on renal function

Patients randomised to VKA got warfarin (~80%) or acenocoumarol (~20%): target INR range 2.0-3.0

INR control got a lot BETTER for those that had been on warfarin prior to the trial that stayed on it

#ESCCongress

What did they find? That patients on rivaroxaban did WORSE:

More death!
More ischaemic strokes!
No diff in major bleeding!

#ESCCongress

You can read an excellent full write-up about the trial by @michaelTCTMD for @TCTMD here:

tctmd.com/news/invictus-…

@mmamas1973 @ShelleyWood2

#ESCCongress

The discussion focuses on how surprised everyone was and what could explain the reason for this unexpected result
Editorial also explores similar issues

But, once again, these results don't surpise me.

Not due to past NOAC data, but due to the RHD hearts...

#ESCCongress
I do a lot of pre-AF ablation TOE to check the LA appendage. What do we do if find thrombus, even in someone on a NOAC?

Often we switch to VKA with a higher target INR than usual (e.g. 2.5-3.0 or 2.5-3.5) and re-TOE in 3 months. LAA almost always looks better

Purely from my TOE experience, I would never have imagined a NOAC could suffice for rheumatic MS. I know it seemed sufficient in the RIVER trial with bioprosthetic MVR, but this is different, I think

#ESCCongress
Bottom line - for rheumatic MS, keep using VKA and not NOAC

Advantage - VKA is cheap. NOAC is expensive
Disadvantage - hassle of regular INR monitoring, logistics of this in LMICs & challenge of a maintaining decent time in therapeutic range (TTR)

Final point

A superb reminder of the importance of following through the evidence chain

The retrospective, observational outcome study *suggested* an association between NOACs and better outcomes

But the subsequent RCT actually showed the complete opposite!

@drjohnm

Observational studies are called that for a reason - one can make *observations*! You can suggest an association but cannot prove causation

Despite PSM in the 2019 study, there must have been residual confounding

So, do the RCT when you can...

Congrats to study authors!