Frontiers in Cardiovascular Medicine (Oct 2021)

Inflammation-Based Scores as a Common Tool for Prognostic Assessment in Heart Failure or Cancer

  • Henrike Arfsten,
  • Anna Cho,
  • Suriya Prausmüller,
  • Georg Spinka,
  • Johannes Novak,
  • Georg Goliasch,
  • Philipp E. Bartko,
  • Markus Raderer,
  • Heinz Gisslinger,
  • Gabriela Kornek,
  • Wolfgang Köstler,
  • Guido Strunk,
  • Matthias Preusser,
  • Christian Hengstenberg,
  • Martin Hülsmann,
  • Noemi Pavo

DOI
https://doi.org/10.3389/fcvm.2021.725903
Journal volume & issue
Vol. 8

Abstract

Read online

Background: Inflammation-based scores are widely tested in cancer and have been evaluated in cardiovascular diseases including heart failure.Objectives: We investigated the impact of established inflammation-based scores on disease severity and survival in patients with stable heart failure with reduced ejection fraction (HFrEF) paralleling results to an intra-institutional cohort of treatment naïve cancer patients.Methods: HFrEF and cancer patients were prospectively enrolled. The neutrophil-to-lymphocyte-ratio (NLR), the monocyte-to-lymphocyte-ratio (MLR), the platelet-to-lymphocyte-ratio (PLR), and the prognostic nutritional index (PNI) at index day were calculated. Association of scores with disease severity and impact on overall survival was determined. Interaction analysis was performed for the different populations.Results: Between 2011 and 2017, a total of 818 patients (443 HFrEF and 375 cancer patients) were enrolled. In HFrEF, there was a strong association between all scores and disease severity reflected by NT-proBNP and NYHA class (p ≤ 0.001 for all). In oncologic patients, association with tumor stage was significant for the PNI only (p = 0.035). In both disease entities, all scores were associated with all-cause mortality (p ≤ 0.014 for all scores). Kaplan–Meier analysis confirmed the discriminatory power of all scores in the HFrEF and the oncologic study population, respectively (log-rank p ≤ 0.026 for all scores). A significant interaction with disease (HFrEF vs. cancer) was observed for PNI (pinteraction = 0.013) or PLR (pinteraction = 0.005), respectively, with higher increase in risk per inflammatory score increment for HFrEF.Conclusion: In crude models, the inflammatory scores NLR, MLR, PLR, and PNI are associated with severity of disease in HFrEF and with survival in HFrEF similarly to cancer patients. For PNI and PLR, the association with increase in risk per increment was even stronger in HFrEF than in malignant disease.

Keywords