Biomarkers to Predict the Response to Cardiac Resynchronization Therapy

Ward Heggermont; Angelo Auricchio; Marc Vanderheyden


Europace. 2019;21(11):1609-1620. 

In This Article

Markers of Inflammation

(High-sensitivity) C-reactive Protein

Heart failure patients display elevated inflammatory markers which correlate with morbidity and mortality.[15] In one study,[16] the predictive power of the most frequently used inflammatory marker, high-sensitivity C-reactive protein (hsCRP) was investigated in 65 HF patients eligible for CRT (46 males, mean age 65 ± 12 years, NYHA III/IV). Levels of both hsCRP and BNP were measured before device implantation. Reverse remodelling, an element of beneficial response to CRT, was defined as >15% reduction in LVESV. In this patient cohort, the hsCRP levels were significantly higher in the non-responders than in the responders (P < 0.01). Moreover, multivariate logistic regression analysis showed a relationship between hsCRP and the incidence of a non-response [odds ratio (OR) 1.499, P = 0.011]. In addition, the strongest predictive factor for cardiac death was an elevated hsCRP (HR 1.337, P = 0.001), with a cut-off of 3 mg/L.[16] In another small pilot study (36 CRT patients) where special focus was attributed to extracellular matrix elements, CRP decreased following successful resynchronization therapy Table 1 and Table 2.[17]

Complement C3A

Another intriguing parameter is the amount of activated complement C3 (C3a) as a marker of chronic inflammatory state. One study investigated prospectively several components of the complement cascade (total C3, C3a, sC5b-9) in 126 HF patients, at baseline and 6 months after CRT implantation. Strikingly, CRT reduced the C3a levels (and sC5b-9 levels), and measuring C3A allowed to predict 5 years of mortality of the patients (C3a level >165 ng/mL aligned with an HR of 4.21, 95% CI 1.65–10.72, P = 0.003), and this effect appeared to be independent of the NT-pro-BNP levels that were measured simultaneously.[18]

Blood Cell-derived Parameters

Since a low lymphocytic count and high neutrophil count in haemograms of patients with HF are associated with a dismal prognosis in chronic HF, these parameters have been investigated in the CRT population. One of the main explanations for these higher neutrophil counts are an increased activation of the innate immune system in HF, and an increased neutrophil life span (viability) in HF patients.[19] In a large observational study by Boros et al.,[20] qualitative blood counts and NT-pro-BNP were analysed, with 2 years of mortality as primary endpoint and reverse remodelling at 6 months (>15% decrease in LVESV) as secondary endpoint. A neutrophil/lymphocyte ratio over 2.95 predicted the absence of reverse remodelling (OR 0.38 with 95% CI 0.17–0.85) and 2 years of mortality, independently of NT-pro-BNP.[20] Another, retrospective, analysis demonstrated the same finding, namely that baseline neutrophil/lymphocyte ratio was significantly higher in non-responders. In this study, patients having a ratio >3.45 had a 12-fold increased risk of CRT non-response.[20] The latter study also investigated platelet to lymphocyte ratio and percentage of lymphocytes, but these values had less predictive value.

Other Inflammatory Markers

Markers that sit on horseback of both inflammation, fibrosis, and remodelling are of considerable interest because they grasp important parts of the pathophysiological alterations in HF. Therefore, interleukin (IL)-6, IL-1α, IL-1β, IL-4 and IL-13, epidermal growth factor, and fibroblast growth factor 2 (FGF-2) were investigated in the multicentre Inflammatory Mediators and Clinical Outcome in Patients with Advanced Heart Failure Receiving CRT (RISK) study.[21] On multivariate analysis of 257 patients, two markers were particularly regulated. Interleukin-13 was significantly associated with the primary outcome, a combination of freedom of HF hospitalizations, death, and decrease in LVESV of >15% at 12 months' of follow-up. Detectable vs. non-detectable IL-13 levels were associated with an OR of 3.79 (95% CI 2.10–6.82, P = 0.0001). Conversely, detectable FGF-2 levels were negatively associated with the primary endpoint (OR 0.31, 95% CI 0.14–0.68; P = 0.004).[21] In the Brouwers study[7] described earlier, only lower tumour necrosis factor (TNF)-α levels were associated with a subjective response to CRT (and not with the objective response), in contrast to any other investigated marker [CRP, IL-6, soluble tumor necrosis factor receptor (sTNFr)1, and sTNFr2]. In a smaller study involving 46 CRT responders and 35 non-responders without any relevant baseline differences, concentrations of IL-6 and TNF-α significantly decreased in the responders to CRT.[22] And finally, levels of cardiotrophin-1 (CT-1), a member of the IL-6 family, significantly decreased in 15 CRT non-responders vs. 29 responders in a pilot study.[23] In a multivariate analysis, baseline CT-1 appeared as an independent predictor of CRT response, at least following echocardiographic standards (>15% decrease of LVESV) (OR 2.7, 95% CI 1.4–4.3; P = 0.01).[23]