OC-05 - D-Dimer, fibrinogen and TEG-MA predict thromboembolism in non-small cell lung cancer - interim results from a prospective cohort study

M Alexander; R Wolfe; D Ball; R Manser; M McManus; B Riedel; B Solomon; D Westerman; K Burbury

doi:10.1016/S0049-3848(16)30122-0

OC-05 - D-Dimer, fibrinogen and TEG-MA predict thromboembolism in non-small cell lung cancer - interim results from a prospective cohort study

Thromb Res. 2016 Apr:140 Suppl 1:S170. doi: 10.1016/S0049-3848(16)30122-0. Epub 2016 Apr 8.

Authors

M Alexander¹, R Wolfe², D Ball³, R Manser³, M McManus³, B Riedel³, B Solomon³, D Westerman³, K Burbury³

Affiliations

¹ Peter MacCallum Cancer Centre, East Melbourne; Monash University, Melbourne; Australia.
² Monash University, Melbourne; Australia.
³ Peter MacCallum Cancer Centre, East Melbourne.

PMID: 27161677
DOI: 10.1016/S0049-3848(16)30122-0

Abstract

Introduction: Cancer associated thromboembolism (TE) is common however the risk is heterogeneous and dynamic.

Aim: To assess the TE risk profile of patients with Non-Small Cell Lung Cancer (NSCLC), though treatment phases, and generate a risk-stratified decision algorithm for appropriate thromboprophylaxis.

Materials and methods: Single centre, prospective observational study, profiling NSCLC patients using clinical-, tumour- and treatment-related risk factors, in conjunction with an extensive thrombogenic biomarker panel, during treatment (surgery, chemotherapy, targeted therapy and/ or radiotherapy) and disease phases. Biomarkers include: FBC, APTT, PT, D-dimer, fibrinogen, FVIIIc, TM, TAT, vWF, prothrombin fragments 1+ 2, fibrin monomers, TEG, microparticles, OHP. Cancer management is per clinician discretion and/or concurrent interventional study protocol. Biomarkers are assessed at baseline; weeks 1, 4 and 12 after commencing cancer treatment; 3 monthly until 12 months.

Results: The interim cohort for analysis included 68 patients, 43 (63%) males, median age 67 years (range 43-67) and with median follow-up 5 months (range 0.2-11). Importantly, 21 patients (15% of patients screened) were ineligible for this study, having presented with TE at diagnosis of NSCLC, while a further 15 patients (15% of study cohort) developed TE while on study. Median time to TE on study was 2.4 months (range 0.1-7). Patients who developed TE demonstrated a biomarker profile indicative of a hypercoagulable state. Khorana score did not adequately stratify or predict TE in this cohort (PPV 20%, NPV 80%), with more than half of patients classified as low or intermediate risk (score 1 (44%), 2 (29%). However, D-dimer ≥1.44mg/L+Fibrinogen ≥4g/L+TEG-MA ≥69mm, as a single measurement at baseline, predicted TE (OR 4.2, p=0.005) and at 4-weeks after commencing cancer treatment (OR 6.7, p=0.005). Predictive power increased further, when considering longitudinal measurements from baseline to 12-weeks after commencing cancer treatment, with OR 14.0 (p<0.001) and PPV 40%, NPV 95%. Inclusion of other Khorana parameters in this model, did not improve predictive performance.

Conclusions: Interim study results reveal a high TE risk among patients with NSCLC. Simple, routine, algorithmic thrombogenic biomarkers demonstrate the capacity to stratify risk and predict TE. Ongoing analyses with the planned larger cohort, expanded biomarker panel and longer follow-up, with longitudinal assessments, are likely to provide greater insight and enhance predicative power. The results will contribute to the development of a simple clinical and biomarker risk stratification tool, to facilitate real-time and dynamic decision-making for appropriate thromboprophylaxis strategies.