Abstract Number: PB1090
Meeting: ISTH 2021 Congress
Theme: Venous Thromboembolism » Cancer Associated Thrombosis
Background: Growth Differentiation Factor 15 (GDF-15), N-terminal pro-B-type natriuretic peptide (NT-proBNP) and high-sensitivity cardiac Troponin T (hs-TnT) are associated with an increased risk of VTE in non-cancer patients, however, the performance of these biomarkers in cancer patients is unknown.
Aims: To assess the performance of GDF-15, NT-proBNP and hs-TnT in predicting VTE in patients with cancer.
Methods: A post-hoc analysis using 1-month plasma samples from patients enrolled in the AVERT trial (a randomized, placebo-controlled, double-blind trial to assess the efficacy and safety of apixaban as primary thromboprophylaxis in ambulatory cancer patients with intermediate to high risk for VTE) to determine if levels of GDF-15, NT-proBNP and hs-TnT are associated with VTE. Logistic regression analysis was used to calculate adjusted odds ratios (OR) across tertiles of these biomarkers. A first model was constructed for all cancer patients, a second excluding patients with brain cancer (justified by lower GDF-15 and NT-proBNP, and the potential influence of the blood-brain barrier), a third for gynecologic cancer. All models were adjusted for age, sex, treatment group, and concomitant antiplatelet therapy [Table 1].
Results: 477 patients were analysed. In Model 1, 2 and 3, participants with highest tertile GDF-levels had adjusted ORs for VTE of 1.66(p=0.199), 3.16(p=0.002), and 4.27(p=0.010) versus the lowest tertile, respectively [Table 2]. For NT-proBNP, levels in the moderate tertiles had a significant association with VTE for patients with all cancers and non-brain cancers [adjusted ORs: 1.84(p=0.011) and 3.28(p=0.033), respectively]. Hs-TnT levels in the moderate tertiles was associated with a decreased risk of VTE compared to the highest tertile as reference in the all cancer and non-brain cancer models.
| Predictors | Venous Thromboembolism (n=477) | |||
| Yes (n=32) | No (n=445) | P-value | ||
| Mean Age, years (SD) | 62.59 (12.0) | 60.87 (15.0) | ||
| Sex (%) | Male Female |
17 (53.1%) 15 (46.9%) |
179 (40.2%) 266 (59.8%) |
0.1520 |
| Treatment (%) | Control Apixaban |
23 (71.9%) 9 (28.1%) |
213 (47.9%) 232 (52.1%) |
0.0087* |
| Antiplatelet Use (%) | No Yes |
27 (76.9%) 5 (15.6%) |
342 (76.9%) 103 (23.2%) |
0.3261 |
| Cancer (%) | Brain Gynecologic Lung Lymphoma Myeloma Pancreas Stomach Breast Other |
4 (12.5%) 8 (25.0%) 0 (0%) 5 (15.6%) 0 (0%) 6 (18.8%) 7 (21.9%) 0 (0%) 2 (6.2%) |
18 (4.0%) 126 (28.3%) 46 (10.3%) 116 (26.0%) 14 (3.1%) 51 (11.4%) 28 (6.3%) 16 (3.6%) 31 (7.0%) |
0.0983 |
| Median GDF-15, pg/mL (IQR) | 2352 (2726) | 1910 (2064) | 0.2036 | |
| Median NT-ProBNP, pg/mL (IQR) | 175 (179) | 111.80 (182.33) | 0.1045 | |
| Median hs-TnT, pg/mL (IQR) | 8.54 (13.08) | 6.10 (7.28) | 0.0661 | |
Descriptive Statistics According to Venous Thromboembolism Events
| Biomarker | Model 1 (n=477) OR (95% CI) |
P-value | Model 2 (n=455) OR (95% CI) |
P-value | Model 3 (n=134) OR (95% CI) |
P-value | |
| GDF-15 Tertiles | Low (ref) (<1415 pg/mL) |
— | — | — | — | — | — |
| Moderate (1415-2581 pg/mL) |
0.87 (0.36-2.06) | 0.747 | 1.62 (0.71-3.72) | 0.253 | 1.41 (0.25-7.86) | 0.695 | |
| High (>2581 pg/mL) |
1.66 (0.77-3.62) | 0.199 | 3.16 (1.52-6.57) | 0.002* | 4.27 (1.42-12.87) | 0.010* | |
| NT-proBNP Tertiles | Low (ref) (<64.8 pg/mL) |
— | — | — | — | — | — |
| Moderate (64.8-189.4 pg/mL) |
1.84 (1.15-2.95) | 0.011* | 3.28 (1.10-9.76) | 0.033* | 2.77 (0.07-112.03) | 0.590 | |
| High (>189.4 pg/mL) |
1.67 (0.74-3.76) | 0.220 | 3.22 (0.64-16.33) | 0.158 | 10.45 (0.45-240.80) | 0.143 | |
| Hs-TnT Tertiles | Low (<4.4 pg/mL) |
0.53 (0.27-1.05) | 0.069 | 0.63 (0.29-1.37) | 0.241 | 0.93 (0.12-7.48) | 0.951 |
| Moderate (4.4 -8.91 pg/mL) |
0.45 (0.26-0.77) | 0.004* | 0.47 (0.24-0.92) | 0.028* | 0.15 (0.02-1.48) | 0.105 | |
| High (ref) (>8.91 pg/mL) |
— | — | — | — | — | — | |
Logistic Regression Models Looking at Associations between Biomarkers and VTE
Conclusions: In this first study to evaluate the predictive performance of GDF-15, proBNP and hs-TnT for VTE in patients with cancer, higher tertile GDF-15 and NT-proBNP predicted increased VTE risk whereas higher hs-TnT predicted decreased VTE risk.
To cite this abstract in AMA style:
Roy D, Wang TF, Carrier M, Mollanji E, Liu P, Wells P. Plasma Biomarkers for Predicting Risk of Venous Thromboembolism (VTE) in Ambulatory Cancer Patients Receiving Chemotherapy [abstract]. Res Pract Thromb Haemost. 2021; 5 (Suppl 2). https://abstracts.isth.org/abstract/plasma-biomarkers-for-predicting-risk-of-venous-thromboembolism-vte-in-ambulatory-cancer-patients-receiving-chemotherapy/. Accessed September 24, 2023.« Back to ISTH 2021 Congress
ISTH Congress Abstracts - https://abstracts.isth.org/abstract/plasma-biomarkers-for-predicting-risk-of-venous-thromboembolism-vte-in-ambulatory-cancer-patients-receiving-chemotherapy/