“HIGH sensitivity Troponins In Patients With Elevated Pro BNP and Acute Heart Failure”. (HIGH-TRIP Trial)

Background: In patients presented to emergency rooms NT-Pro BNP essay is overly sensitive test to rule out heart failure but less specic in predicting outcomes in follow-ups, in this study we ought to nd the added value of HS-Troponin I, in patients presented acutely with heart failure and its impact on mortality when Pro BNP is highly elevated. Methods: Prospective cohort study, inclusion criteria were age above 18 and clearly positive NT Pro BNP > 1000 pg/ml, with 12 months follow up period, primary end point was mortality from heart failure, secondary endpoint was need for rehospitalization.


Conclusion:
Adding (Hs-cTroponin I) assay to the panel of laboratory testing, in patients presented to ER with acute heart failure and with high Pro-BNP>1000, may further predicts mortality and rehospitalization rate.

Background
Acute cardiac pulmonary edema secondary to heart failure carries a bad prognosis. Beta natriuretic peptide (BNP) and N-terminal Pro BNP are the gold standard biomarkers used for diagnosis and prognosis, more novel biomarkers like mid-regional pro adrenomedullin (MR-proADM) mid-regional pro atrial natriuretic peptide (MR-proANP), highly sensitive troponins Hs-cTn, soluble Suppression of tumorigenicity 2 (sST2), growth differentiation factor (GDF)-15 and Galectin-3, may have a role to play in determining prognosis far and beyond the one established by natriuretic peptides, but their potential role in clinical care of patients with heart failure is yet to be fully determined (1).
There are many studies that looked at the utility of using pro-BNP, the natural peptides produced by atria, due to stretching effect of elevated cavitary pressure, and its value in diagnosing pulmonary edema. It can discriminate cardiac from other causes of lung in ltrates like acute respiratory syndrome and pneumonia. Albeit very well studied (2,3,4,5,6,78), BNP and Pro BNP still have some limitations particularly when they are used in follow ups as results of assays may be affected by several factors related to age, gender, body mass, renal function and exercise, all of which may limit its utility for accurately stratifying some patients with acute heart failure (9, 10, 11, 12, 13&14).
(Hs-cTn) High sensitivity cardiac troponin has been suggested to play a role in predicting early onset of heart failure in asymptomatic patients (15,16) and as a marker of myocardial injury in patients with the rst ever presentation of heart failure, based on the Biomarker for Cardiovascular Risk Assessment in Europe consortium (BIOMAReCare) Hs-cTn shows its independence for the prognosis of HF (17,18,19,20).
In this study we ought to nd the added value of Hs-cTn I in patients presented acutely to the emergency room with dyspnea and highly elevated Pro BNP and its impact on predicting death and re-hospitalization over 12-month time.

Methods
Cohort prospective study with inclusion criteria of adults above the age of 18, with suspected acute pulmonary edema due to heart failure and elevated NT-Pro BNP >1000 pg/ml. Study was approved by the IRB of Dr Sulaiman Alhabib Medical Group holding approval number HAP-01-R-082, all experiments were performed in accordance with relevant guidelines and regulations and was Health Insurance Portability and Accountability Act (HIPPA complaint). informed consent was obtained from all subjects and/or their legal guardian(s). After recruitment, Hs-cTn plasma level was assessed in addition to other laboratory investigations including creatinine, as well as hemoglobin level. All patients underwent chest Xray at presentation. X-rays were read independently by expert radiologists and patient were subclassi ed into pulmonary edema and non-pulmonary edema based on the reports' results and were followed for 12 months. Mean of all variances were calculated and analysis of variance (ANOVA) was conducted for noncontinuous variables, while (t-test was used for continuous non categorical variables, using 95% power level, looking for p<0.05 of signi cance. During follow up patient's data were collected including echocardiographic ndings, additional special intervention including aggressive medical therapy and revascularization with Percutaneous intervention (PCI) or Bypass surgery (CABG), primary end point was mortality from heart failure and secondary endpoint was need for rehospitalization. Patients who lost follow up were censored and results were not counted in the nal analysis. (Figure1: Flowchart of the study) explains the study design. Hs-cTns were measured using (Abbott Laboratories; Chicago, IL) with reference range values of (0.004-0.028 ng/ml) and N-terminal of the prohormone brain natriuretic peptide NT (Pro BNP) was assessed via Elisa immunoassay with 460 pg/ml been the cut value of elevated hormones, we used 1000 pg/ml to avoid borderline positive tests. We did not look at repeated levels of troponins or pro BNP in subsequent presentations.

Results
After 12 months follow up & a mean of 180 days, 95 patients were recruited of which 64 were with nonovert pulmonary edema (non-PE) and 31 with pulmonary edema (PE), mean age was 78+/-6 and 68+/-6, males were 51 & 50% on each group. Mean Pro BNP levels were 6148, 5927 pg/ml while Hs-cTn were 19.27 and 0.17ng/ml with mean Body mass index (BMI) of 35 and 29.8, respectively. Other risk factors and clinical ndings are depicted on (table 1), of note there were more cases of atrial brillation in the non-overt pulmonary edema group compared to the overt pulmonary edema 41vs 19%. Mean creatinine and hemoglobin blood levels were comparable between both studied groups.
Echocardiographic ndings revealed no signi cant difference in overall ejection fraction with mean values around 48 to 47%, similarly the left atrial size, early diastolic to late diastolic mitral in ow doppler velocity (E/A) ratio, Tricuspid Regurgitation (TR) jet velocity were equivalent between both groups, two exceptions were observed including high Early mitral in ow velocity to early mitral tissue doppler annular velocity E/e' of 30 compared to 7 in favor of PE group and the regionality on resting echo was higher among patients in PE group in at least 50% of cases vs only 26% in the non PE (table 2).
For endpoints, we identi ed 4 mortalities related to heart failure (cardiogenic shock) in the PE compared to 1 in non-PE cohorts, for secondary endpoint 12 cases were hospitalized in the span of 365 days in PE compared to only 7 in the non-PE. The odd ratios were 8.5 and 4.8m, p= 0.003 and 0.008 respectively (Table 3), in the COX proportional hazard regression covariate analysis, 5 (covariates) factors were suspected to predict poor outcomes including Pro BNP, Troponin, E/e' ratio, regionalities and overt pulmonary edema on Xray at the time of presentation, Hs-cTn was the only signi cant predictor, Hs-cTn is more sensitive and speci c compared to Pro BNP in layering the risk of death and accurately predicting poor outcomes, receiver operator curves were obtained for both and depicted on ( Figure 2) and (3). with omnibus multivariate logistic regression statistic showed signi cant p =0.001, cumulative survival and hazard risk with mean (Hs-cTnI) of 17 has been depicted in (Figure4). Furthermore, need for revascularization was assessed, with 3 vs 7 patients underwent Coronary artery bypass surgery (CABG) and 8 vs 12 patient had Percutaneous Coronary Intervention (PCI) in PE and Non-PE group, respectively. Risk of stroke were similar over all between PE and non-PE groups' but higher in the subgroup of patient who underwent CABG at 2% compared to overall of 1%. Patients were placed on good medical therapy throughout the study with excellent rate of compliance of more than 85% of cases and were followed in heart function clinic.

Discussion
Although both were studied before the added value of Hs-cTn to BNPs is an area that is still under scrutiny (21,22,23). The study shed lights on the value of adding one to another in further predicting hard endpoints of heart failure related death and re-hospitalization. There are no "perfect" tests but in this scenario Hs-cTns were "perfect" with sensitivity of 96%. The depicted ROC (receiver operator curves) speaks to this effect very illustratively.
We believe the ndings of this trial is unique and may be practice changing for treating patients with acute heart failure. We have selected extremely high-risk patients with quite elevated level of pro Bnp >1000 and we further stratify them using Hs-cTns and radiological evidence of pulmonary edema, both biomarkers are measured in the laboratories and their blood levels are pathophysiologically explained.
BNPs are produced due to stretching effect of the myocardial tissue while high sensitivity troponins are released in blood stream in response to direct damage to the myocardium, (24,25). Translating the laboratories nding to clinical practice was in keeping with the clinical outcomes of patients in this trial, as myocardial damage (measured by high sensitivity Troponins) is more of a surrogate marker compared to the stretching effect causing (Pro BNP elevation).
Despite the novel nding of the trial, there are quite few limitations of the study that need to be highlighted, including small numbers of patients in both groups leading to wide con dence intervals and reducing the power of the study, the selection bias related to open label protocol, and potential lab errors related to sample drawn.
The study seems to touch upon major factors of increasing risk of deaths in patients with heart failure including risk factors of ischemic heart disease, reduced ejection fraction, elevated lling pressure by echo criteria and radiological evidence of pulmonary edema, however the study was not powered to look at the impact of all these factors in the small size sample of patients recruited, making it very di cult to compare its utility in addition to Hs-cTns. We did not repeat cardiac troponin level in post recruitment presentations to avoid dilutional effect of intervention on blood level of pro BNPs and troponins.
Focusing on predicting outcomes is important aspects of proper management of patient with heart failure as outlined in most recent guidelines, where class A is de ned as patient at risk of heart failure but not yet diagnosed to have heart failure (26), adding Hs-Ctn to the paradigm of heart failure management may be of value to tailor more aggressive surveillance and more strict management protocol that will help improve survival and outcomes.
The trial served its purpose very well, the drop-out rate was 5%, and results were clinically signi cant even with the inherited limitations we have alluded to. Further study with large numbers of patients is needed to address the same question but in the setup of other factors or other novel markers, the later may be better in re ning the risk assessment tool of patient with rst time presentation of heart failure.

Conclusions
In this small cohort of patient with acute pulmonary edema due to heart failure, high sensitivity cardiac troponins were successful in predicting poor outcomes related to death and hospitalization, High HS-TROPS has odd ratio of 4.3 and 8.5 in predicting both respectively despite equal mean level of PRO BNP and ejection fraction.    Figure 1 Flowchart of study design Receiver operator Curve for Hs-cTn X-Axis is sensitivity or True positive Y-Axis is 1-spec city or True negative rate Blue line: is test performance with area under the curve of 97% Green line: is when can't differentiate true positive from true negative as chances are 50% for both.

Figure 3
Receiver Operator Curve of Pro BNP X-Axis is sensitivity or True positive Y-Axis is 1-spec city or True negative rate Blue line: is test performance with area under the curve of 62% Green line: is when can't differentiate true positive from true negative as chances are 50% for both.