@stuchman1 @Amyloid_Planet @vsanchorawala There was actually a recent article in Haematologica which showed that Mayo 2004+IIIb may be the most accurate staging system. It'll be interesting to see if this conclusion bears out, since then our staging system/Mayo 2004+IIIb would be the best. .

Thank you for having us @Amyloid_Planet! #amyloidosisJC

@vsanchorawala Here is the treatment regimen breakdown for our "complementary cohort" that shows the use of many contemporary treatments including stem cell transplant that probably explain the significant increases in survival @amyloidosisJC

@Amyloid_Planet @vsanchorawala So we thought about this; however, our 1st cohort was from March 2016 - Sep 2016 and we were doing the data analysis in early-mid 2018, which means that we would have had at most 24 months of follow-up, which may not have shown differences between Stage I/Stage II @amyloidosisJC

@Amyloid_Planet @vsanchorawala We would have loved to be able to do this, but unfortunately we couldn't since nobody had NT-proBNP drawn at our center prior to 2016. Hence why we had to use the two separate cohorts at different time frames. #amyloidosisJC

@a_staron @vsanchorawala It's actually years on the x-axis #amyloidosisJC

@Amyloid_Planet @stuchman1 Maybe the way to think about it (@vsanchorawala correct me if I'm wrong), for survival, it doesn't matter *how* the cardiac biomarkers are -- whether through cardiac involvement or ESRD. Just the fact that they are elevated means worse survival. #amyloidosisJC

@Amyloid_Planet @stuchman1 The specific ROC curves we did were only for predicting cardiac involvement. Our survival curves in our complementary cohort were agnostic to CKD stage (and we didn't remove any patients with ESRD from the analysis), which is similar to all Mayo staging systems. #amyloidosisJC

So in the end, here's the way to stage patients with BNP with the resulting survival curves seen in @vsanchorawala recent post. #amyloidosisJC

To determine stage IIIb, we did ROC analysis for 1 year survival on the patients who were in stage III (i.e. TnI > 0.1 and BNP >81), and we found a "IIIb" BNP cutoff of >700 pg/mL #amyloidosisJC

So in our validation cohort, 151 patients were assigned to stage I, 259 patients were assigned to stage II, and 88 patients wereassigned to stage III, and 94 were assigned to stage IIIb. #amyloidosisJC

So once we had shown that a BNP >81 cutoff closely mimic'd an NT-proBNP cutoff of >332, we had to validate the system, which we did using a cohort of 1073 patients (592 of which had BNP drawn and could be used) seen for 1st visits between 2004-2014 #amyloidosisJC

Interestingly, NT-proBNP had better specificity in patients with NO CKD and then in all comers there was no difference. This would suggest towards BNP's superiority in CKD IV-V; however, we didn't have enough patients to prove this. #amyloidosisJC

@stuchman1 We did after 2016 at our center -- primarily because there wasn't any data out there to support BNP use and so we needed to use the gold standard. @vsanchorawala would be able to explain more. #amyloidosisJC

We also looked at what BNP threshold would best predict cardiac involvement (based on a combination of endomyocardial biopsy, MRI, and echo criteria) and it turns out that a BNP of 81 pg/mL also best predicted cardiac involvment (BNP is left, NT-proBNP is right) #amyloidosisJC

Out of the 250 patients, 249 had both biomarkers drawn and we derived a BNP cutoff (which turned out to be 81pg/mL) that most closely mimic'd the Mayo 2004 system. It turns out, by using a cutoff of 81, just over 89% of people will be given the same stage. #amyloidosisJC

@stuchman1 So they're created in an equimolar fashion from pre-proBNP and released in a 1:1 ratio. NT-proBNP is ONLY cleared via the kidneys whereas BNP is also cleared in the bloodstream.