heart failure: check BNP measurements regularly

 

A recent study found that repeatedly tracking NT-proBNP levels in patients with heart failure with reduced ejection fraction was a much stronger predictor of developing subsequent clinical heart failure than using just a baseline value (see heart failure repeated BNP predicts HF exacerb JACC-HF2023 in dropbox or doi.org/10.1016/j.jchf.2023.11.007)

 

Details:

-- 894 study participants who had chronic heart failure (HF) with reduced ejection fraction were included from 45 outpatient sites in the US and Canada, in the GUIDE-IT trial. The current study is a secondary analysis of this trial

-- the GUIDE-IT trial tested the impact of titrating guideline-directed medical therapy (GDMT) using a target NT-proBNP<1000 pg/mL for those with a baseline NT-proBNP >2000 pg/mL, versus usual care

    – this GUIDE-IT trial was stopped early because of futility/no difference between the guided therapy vs placebo, with each having 37% of patients with a composite of HF or cardiovascular mortality

-- patients had repeated NT-proBNP levels measured every three months over a two-year period

-- mean age 63, 32% female, 55% white/37% Black, BMI 29

-- ejection fraction 24%, NYHA functional class II in 51%/ class III in 40%

-- baseline laboratory values: NT-proBNP 2607 pg/mL, serum sodium 139, serum potassium 4.3, eGFR 52 by creatinine

-- comorbidities: ischemic heart disease 51%, diabetes 46%, implantable cardioverter-defibrillator 44%, OSA 23%, atrial fibrillation 16%, depression treated with medications 16%

-- site of studies: academic 55%, community 45%; 85% in the US/15% in Canada

-- medications used: beta blocker 96%, ACE/ARB 81%, MRA 50%, loop diuretic 94%

 

-- Primary endpoint: cardiovascular death or heart failure hospitalization, comparing the prognostic value of a baseline NT-proBNP vs repeated NT-proBNP evaluations every three months

    -- this evaluation assessed how patients would have been correctly reclassified using the three-month or six-month NT-proBNP values, versus using the single baseline value, in terms of the actual 12-month clinical outcomes of cardiac death or heart failure hospitalization

    -- the results of cardiovascular death or heart failure hospitalization were adjusted for the usual baseline demographics, medical conditions, and NYHA functional class

 

Results:

-- HF hospitalizations or cardiovascular mortality:

    -- 328 (37%) of the patients had a HF hospitalization or died from cardiovascular causes

    -- baseline NT-proBNP levels were higher among the individuals who subsequently had an primary event: 3377 pg/mL (1876-6524) versus 2311 pg/mL (1264-4705), p<0.001

    -- using just the baseline levels, there was a 17% increased risk of these outcomes for a doubling of the NT-proBNP values, HR 1.17 (1.08-1.28), p=0.0003

    -- for those 621 patients who did not have a primary outcome for the first 3 months, then had NT-proBNP measurement at 3 months, using this longitudinal assessment model incorporating the 3-month NT-proBNP value improved the outcome prediction: there was a 66% increased risk for these primary endpoints for a doubling of the NT-proBNP values, HR 1.66 (1.50-1.84), p<0.0001

        --ie, there was much more accurate prognostic information by subsequent assessment of NT-proBNP measurement at 3 months

 

-- All-cause mortality:

    -- assessing baseline NT-proBNP only: 59% increased risk, HR 1.59 (1.38-1.83), p<0.0001

    -- assessing follow-up NT-proBNP measurements after the baseline measurement: 75% increased (1.53-2.01), p<0.0001, but this difference was not statistically significant

 

-- HF hospitalizations:

    -- assessing baseline NT-proBNP only: 16% increased risk, HR 1.16 (1.06-1.27), p<0.0001

    -- assessing follow-up NT-proBNP measurement at 3 months after the baseline measurement: 64% increase, HR 1.64 (1.47-1.84), p<0.0001

        -- ie, much greater prognostic accuracy by assessing the 3-month NT-proBNP measurements

 -- This graphic indicates how patients fared clinically, from evaluating just their baseline 12-month risk category vs having the additional 3-month NT-proBNP measurements: the left side representing the number of patients who either did or did not develop heart failure requiring hospitalization or cardiovascular death 12 months later based solely on their baseline NT-proBNP; the right side was for those having or not having these clinical outcomes, with specific information on how the initial group would have been reclassified if there were 3-month NT-proBNP results available. 

 

-- so, a single measurement at three months later (right side) led to a dramatic reclassification for these events occurring at the 12 month time period:

    -- among the 190 participants who had a subsequent heart failure hospitalization or cardiovascular death at the 12-month mark, utilizing the 3-month NT-proBNP would have had them classified quite differently: 

        -- 96 (34%) were correctly classified as being in the high risk category by just the baseline NT-proBNP

        -- 30 (16%)  were incorrectly classified by the baseline value as being high risk/ would have been correctly reclassified to a lower risk category with the 3-month NT-proBNP results

        --192 (51%) were incorrectly classified by the baseline value as being low risk/would have been correctly reclassified to a higher risk category

            – ie, 67% of those who had an event were reclassified appropriately if the 3-month NT-proBNP were performed

     -- among the 431 participants who did not have a subsequent heart failure hospitalization or cardiovascular death at the 12-month mark, utilizing the subsequent 3-month NT-proBNP would have had them classified, by risk for a subsequent event:

        -- 176 (40%) were correctly classified to a lower risk category by just the baseline NT-proBNP

        -- 192 (45%) were incorrectly classified by the baseline value as being high risk/would have been correctly reclassified to a lower risk category with the 3-month NT-proBNP results

        -- 63 (15%) were incorrectly classified by the baseline value as being low risk/would have been correctly reclassified to a higher risk category

            – ie, 60% of those who did not have an event were reclassified appropriately if the 3-month NT-proBNP were performed

-- and, the 6-month NT-proBNP levels were significantly more accurate in predicting the 12-month clinical outcomes than the three-month levels

-- also the NT-proBNP levels had a positive slope about 200 days before one of these outcomes, and this slope was rising more steeply closer to a clinical event

 

Commentary:

-- BNP measurements have become an important part of heart failure management, and, when persistently high, they predict future HF deterioration and can lead to appropriate GDMT (guideline-directed medical therapy) that is associated with impressive clinical improvement. However, many patients have only a baseline BNP level done, and there are not a lot of data supporting using serial measurements (and some prior studies have had mixed results, with one meta-analysis for monitoring BNP in heart failure patients finding no association with benefit over usual care, and this was independent of whether the BNP or NT-proBNP were measured: see chf bnp guided rx vs UC metaanal IntJCardiol2018 in dropbox, or https://www.internationaljournalofcardiology.com/article/S0167-5273(18)31346-9/fulltext; although another one did find benefit: https://academic.oup.com/eurheartj/article/35/23/1559/459004?login=false  ). In the initial GUIDE-IT study report (a negative study) they only included the patients who had NT-proBNP levels that were >2000, or BNP>400 pg/mL within the prior 30 days or history of a HF event, in both the control and the guideline-based medical treatment by the meds of the time (2013-2016), with the study ending early because of the assessment of intervention "futility" (see https://jamanetwork.com/journals/jama/fullarticle/2649188 ). this current study suggests that in the same cohort in the same study, there was benefit for a longer study with repeated NT-proBNP  measurements.

-- In sum, this study did show strong associations for the following:

    -- the three-month post-baseline assessment of NT-proBNP dramatically changed the predicted value of the baseline NT-proBNP in terms of subsequent heart failure hospitalizations or cardiovascular deaths either happening or not

    -- the six-month assessment of NT-proBNP added significantly more predictive value

    -- the clinical events assessed at 12 months were preceded by increased NT-proBNP levels by about 200 days, with increasing benefit when the increased NT-proBNP was closer to an event

 

-- as a support for these conclusions, a recent study of repetitive BNP measurements in patients with stage 3 to 5 CKD, was associated with decreased heart failure hospitalizations and all-cause mortality: https://gmodestmedblogs.blogspot.com/2023/12/routine-bnp-assessment-helpful-for.html

 

Limitations:

-- studies of heart failure treatment are a bit of a moving target, given the more recent incorporation of some great newer medications, including sacubitril/valsartan (especially in those with a reduced ejection fraction), SGLT-2 inhibitors, GLP-1 agonists (the latter having effectiveness in those with preserved ejection fraction)

    -- GDMT in general is pretty straightforward in an outpatient setting by nonspecialists. The addition of sacubitril-valsartan has been a huge benefit in those with reduced ejection fraction and is usually not difficult in the outpatient setting (https://gmodestmedblogs.blogspot.com/2022/05/heart-failure-outpatient-initiation-of.html ). In addition the availability of SGLT-2 inhibitors, as well as GLP-1 agonists for those with preserved ejection fractions, have also made significant improvements in clinical outcomes

-- this was an observational retrospective assessment of an ongoing prior trial, so therefore conclusions cannot be assumed to be causal, just that there was an association between repeated NT-proBNP measurements and some 12-month clinical outcomes

-- the endpoints of this trial were quite severe: heart failure hospitalizations and cardiovascular deaths. There clearly are significant quality-of-life impairments associated with patients having less severe heart failure. It may well be that aggressively treating significant increases in the NT-proBNP would additionally have important quality-of-life benefits with periodic assessment, as well as these severe outcomes

-- also, they only included those with NYHA class II or III heart failure. Would there be benefit for those with class I or class IV???

-- this study did not test whether the sequential NT-proBNP measurements led to improved clinical outcomes by initiating GDMT earlier, only that the prognostic value of finding subsequent changes in NT-proBNP did reclassify most patients into different risk categories for developing HF (though this seems very likely, given many other studies finding that treating patients with elevated BNPs is clinically beneficial) 

-- we do not have clear information as to the appropriate interval of testing for NT-proBNP. They chose every three months, which seems reasonable based on their finding that the six-month check was significantly more predictive of clinical outcomes than the three-month check, and the lead time from increased NT-proBNP to actual clinical events was 200 days. But this 3-month interval was arbitrary (not tested in RCT) and therefore not evidence-based. Perhaps checking every 2 months would be better (clinical adverse events tended to be closer to the increase in NT-proBNP)? also for how long should we be doing this testing?

-- patients enrolled in the study were treated by heart failure specialists in the setting of a clinical trial, quite different from outpatients that we see in primary care. However, at least in my experience, GDMT has been pretty easy to implement. And, we in primary care are probably in a better position to do regular testing for NT-proBNP levels on a three-month basis, as per the study, since we tend to see sicker patients more often than cardiologists do, given that HF patients typically have many other (and often associated) medical conditions, though, of course, some patients may need more aggressive cardiologic interventions

-- this study was limited to patients with heart failure and reduced ejection fraction. The treatment for those with heart failure with preserved ejection fraction is quite different (the pathophysiology is also quite different: https://gmodestmedblogs.blogspot.com/2023/10/heart-failure-preserved-in-obese.html , which is why GLP-1 agonists work so well). It is therefore a bit of a stretch to apply their conclusion to those with preserved ejection fraction, though it does make intuitive sense

-- we do not have information on those with regular BNP being measured (vs the NT-proBNP). is BNP as good? is it better (less variability by renal function)? is it worse (more variability by which lab measures it)?

 

So, this is a study suggesting a further refinement to heart failure management in those with reduced ejection fraction:

-- guideline-directed medical therapy (GDMT) for patients with heart failure has evolved dramatically in the last several years, with a much wider array of very effective medications

-- the algorithms to treat heart failure are reasonably straightforward, and in general they are implementable in an outpatient primary care setting for most patients

-- it certainly makes intuitive sense to check serial measurements of BNPs. Circumstances/clinical conditions do change. and this can affect either increased or decreased likelihood of a severe cardiovascular event that would be anticipated from subsequent values

-- we in primary care are in a great position to monitor heart failure symptoms as well as periodic assessments of NT-proBNP levels, perhaps every three months, since we tend to see patients more often than specialists, and we are the ones whom patients call when they are developing early symptoms

 

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs:

 

 go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

 

if you would like to see the articles in this blog, please email me. 

 

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list