Skip to main content
  • Antiangiogenic Isoforms of Hepatocyte Growth Factor (HGF) Confound Its Detection as a Biomarker for Risk of Stroke

    Final Number:

    Nestor R. Gonzalez MD; Hao Jiang MD; Michael Schiraldi MD, PhD; Raymond Liou BS; Florian Kurth MD, PhD.

    Study Design:
    Laboratory Investigation

    Subject Category:
    Basic Science

    Meeting: AANS/CNS Cerebrovascular Section 2018 Annual Meeting

    Introduction: Hepatocyte growth factor (HGF) has been associated with risk of stroke in several studies including the Women’s Health Initiative, the MESA (Multi-Ethnic Study of Atherosclerosis) and the ANFIS (Angiogenic Factors in Intracranial Stenosis). HGF is an alpha-beta heterodimer, which is only proangiogenic in this form. However, naturally occurring and artificially produced fragments of the alpha subunit are antiangiogenic. The precise location of the epitope targeted by the antibodies of the commercially available immunoassays is unknown. Therefore, there is a possibility of cross-reactivity or interference caused by antiangiogenic isoforms of HGF, which has never been assessed. We tested the hypothesis that the highly antiangiogenic fragment of HGF, NK4, is detected by immunoassays as HGF.

    Methods: Commercially available ELISA plates for HGF were used to quantify sequential dilutions of decreasing known concentrations of HGF and equimolar NK4. We compared the measured results (HGF-m) with the actual HGF and NK4 (HGF-a and NK4-a) concentrations. The affinity of the HGF ELISA antibodies to HGF and NK4 was evaluated in terms of correlation between HGF-m and HGF-a and NK4-a, and the reliability of the tests was tested with standard Cronbach’s alpha. We calculated the sensitivity, specificity, and accuracy of HGF-m to detect NK4-a and HGF-a.

    Results: HGF-m on ELISA detected NK4-a levels “as HGF” at all concentrations tested (Range 0.1 to 10,000 pg/mL) with a Spearman’s rho of 0.98 (p<0.001). This was indistinguishable from the HGF-a detection (Figure 1), with a very high reliability Cronbach’s alpha of 0.83. Immunoassays had reduced sensitivity (0.26), specificity (0.56), and global accuracy (0.53) to differentiate HGF from NK4.

    Conclusions: The detection of anti-angiogenic NK4 with immunoassays for HGF opens the possibility that elevated concentrations of measured HGF associated with stroke may be caused by circulating antiangiogenic splicing variants of HGF such as NK fragments. Further evaluation of the role of these antiangiogenic isoforms in stroke is necessary.

    Patient Care: The identification of the potential antiangiogenic isoforms of angiogenic factors, such as HGF associated to stroke can become targets for therapeutic intervention in patients at risk

    Learning Objectives: To discuss the potential role of antiangiogenic HGF isoforms in stroke


We use cookies to improve the performance of our site, to analyze the traffic to our site, and to personalize your experience of the site. You can control cookies through your browser settings. Please find more information on the cookies used on our site. Privacy Policy