Rivaroxaban compared to no treatment in ER-negative stage I–III early breast cancer patients (the TIP Trial): study protocol for a phase II preoperative window-of-opportunity study design randomised controlled trial

Breast cancer in women is associated with a four-fold increased risk of venous thromboembolism (VTE) compared to age-matched women without cancer [1, 2]. Chemotherapy-induced VTE occurs in up to 7% of early breast cancer and 17% of metastatic breast cancer patients [3]. Breast cancer patients who develop VTE have a worse survival, which cannot solely be accounted for by VTE-induced mortality [4, 5]. It is thought to in part be due to a promotion of cancer growth and metastasis by coagulation factors.

Tissue Factor (TF), a transmembrane glycoprotein, is the initiator of the extrinsic clotting cascade. TF binds to plasma factor VII/VIIa (FVII/VIIa) to initiate coagulation via activation of Factor X (FX). This leads to the generation of thrombin, fibrin deposition and clot formation, as well as platelet activation (Fig. 1). In addition to its haemostatic role, TF activates cell signalling through TF:FVIIa and TF:FVIIa:FXa cleaving of the G-coupled protease-activated receptor (PARs) 1 and 2 complexes.

TF directly promotes tumour growth, migration and angiogenesis via activation of the PAR-2 receptor [7‐11]. Thrombin can also modify tumour cell behaviour, directly through the activation of PAR receptors, indirectly by acting on stromal cells, or by the generation of fibrin matrices [12‐14]. Thrombin promotes tumour growth in vitro [15] and in vivo [16‐18], enhances tumour adhesion to platelets [16, 17, 19] and endothelial cells [20] in vitro and increases experimental (tail-vein injection) metastasis [16, 17, 19, 21]. Thrombin also increases tumour cell motility [22]. TF [23] and thrombin [24] promote neo-angiogenesis via upregulation of VEGF (a potent angiogenic factor) in cancer cells through the PI3K/Akt pathway [25]. TF also downregulates anti-angiogenic factors such as thrombospondin-1 and thrombospondin-2 [26], while thrombin upregulates matrix metalloproteinase-2 [27]. TF and thrombin also activate angiogenesis by clotting-dependent mechanisms through production (via thrombin) of polymerised fibrin which provides a scaffold around tumours to support and stimulate endothelial cell proliferation [28]. In clinical studies, TF expression correlates with VEGF expression and markers of angiogenesis [29], while prothrombin fragments (produced when prothrombin is converted to active thrombin) closely associate with areas of neo-angiogenesis at the tumour-stroma interface [30, 31].

Cancer cells express high levels of TF, thrombin and PARs [32, 33]. Breast cancer expression of TF has been reported to be an independent predictor of overall survival [34]. Plasma D-dimer (fibrin degradation by-product) is increased in breast cancer and correlates with tumour stage, poor prognosis molecular phenotypes oestrogen receptor (ER) negative, human epidermal growth factor receptor 2 (HER-2) positive [35, 36], circulating tumour cells [37, 38] and reduced survival [39]. D-dimer also has potential as a pharmacodynamic biomarker for chemotherapy response [40].

In a prospective cohort study of 248 early breast cancer patients, we found that tumour fibroblast expression of TF was increased in patients with primary breast cancer as compared to both ductal carcinoma in situ (DCIS) and normal breast. We also found that fibroblast TF, thrombin, PAR-1 and PAR-2 expression were increased in patients with ER-negative, HER-2-positive and high Ki67 tumours [41].

Inhibition of this coagulation cascade could therefore be a therapeutic option in breast cancer patients. A Cochrane review in 2011 reported that parenteral anticoagulation (with either unfractionated heparin or low molecular weight heparin) for patients with cancer, who have no therapeutic or prophylactic indication for anticoagulation, resulted in a statistically and clinically significant reduction in mortality at 24 months. This demonstrates a potential anti-cancer effect of anticoagulants in patients with better prognosis metastatic disease [42].

Direct oral anticoagulants (DOACs) inhibit thrombin (e.g. dabigatran etexilate (Pradaxa®)) or its activating enzyme Factor Xa (e.g. Rivaroxaban). DOACs have come into regular clinical use in the last decade for the prevention of VTE in orthopaedic surgery and of stroke in non-valvular atrial fibrillation [43, 44]. DOACs are also used for the treatment of primary or recurrent VTE. DOACs target a specific single enzyme within the extrinsic clotting pathway unlike the more widespread effects of heparins and vitamin K antagonists (Warfarin). A meta-analysis of randomised controlled trials of VTE-treatment with DOACs in cancer patients has indicated they are as safe and effective as conventional treatments [45].

Rivaroxaban was the first orally active direct Factor Xa inhibitor and, through inhibition of the Tissue Factor-Factor VIIa-Factor Xa complex, inhibits the conversion of prothrombin to thrombin [46]. Rivaroxaban is well absorbed from the gut and maximum inhibition of Factor Xa occurs 1 to 4 h after administration [47, 48]. Rivaroxaban has shown comparable safety to heparin for thromboprophylaxis after total hip replacements, with the convenience of oral dosing and reduced coagulation monitoring [49]. Rivaroxaban has the potential to downregulate the TF-FVIIa-FXa complex activation of PAR-2 and the subsequent tumour cell migration and production of pro-angiogenic and immune-modulating cytokines, chemokines and growth factors [10]. Rivaroxaban was chosen as the DOAC for this window-of-opportunity study as it inhibits the coagulation cascade at its origin and has a good safety profile. We have also found that Rivaroxaban inhibits cancer stem cell (CSC) activity in the in vitro functional CSC assay of mammosphere formation [50].

In summary, in vitro and in vivo evidence supports a potential inhibitory effect of anticoagulants on cancer growth through proliferative, cancer stem cell, angiogenic and metastatic processes. ER-negative early breast cancer patients have a relatively poorer prognosis and higher stage at diagnosis with an associated reduced disease-free survival. ER-negative early breast cancer patients also have fewer options for adjuvant therapy. Our previous data supports the hypothesis that the extrinsic clotting pathway is more upregulated in the ER-negative breast cancer subgroup as compared to ER-positive breast cancer, and thereby provides a potential therapeutic strategy.

To test a novel therapy in the adjuvant breast cancer setting would require thousands of patients, long follow-up and great expense. The alternative is the more novel ‘window-of-opportunity’ clinical trial design where drugs are tested over a short period, in the preoperative setting. Biomarkers are used as early markers for response and long-term outcome. Ki67 (tumour proliferation marker) is recognised as such a marker in trials of endocrine therapy. Short-term neoadjuvant studies have demonstrated that only 2 weeks of preoperative endocrine therapy induces changes in Ki67 which predicts for treatment benefit and long-term survival outcome [51]. By using biomarkers, we may be able to demonstrate the potential effectiveness of an anticoagulant in the clinical setting to support future larger adjuvant trials.

The TIP Trial will assess the anti-proliferative and other anti-cancer progression mechanisms of Rivaroxaban in ER-negative early breast cancer patients.

This study protocol has been reported in accordance with the Standard Protocol Items: Recommendations for Clinical Interventional Trials (SPIRIT) guidelines (Additional file 1).

Ki67 is recognised as a marker of tumour proliferation in trials of endocrine therapy. Short-term neoadjuvant studies have demonstrated that only 2 weeks of preoperative endocrine therapy induces changes in Ki67 which predicts for treatment benefit and long-term survival outcome [51]. By using biomarkers, we will be able to demonstrate the effectiveness of Rivaroxaban in the clinical setting within 5 years, thereby providing an improved clinical trials service to the patients recruited. However, the measurement of Ki67 has its challenges as there is substantial heterogeneity and variable levels of validity in methods of assessment [57]. It is agreed that we will use the immunohistochemistry (IHC) method for determining Ki67 percentages. The IHC staining and assessment of Ki67 will be performed on all trial samples in one batch upon trial completion in an attempt to minimise variability and increase accuracy.

The collection of fresh tumour tissue for mammosphere analysis will not always be feasible, particularly in patients with smaller tumours or who undergo a research biopsy. Mammosphere analysis is therefore a secondary and exploratory outcome of the TIP trial. However, if successful, it may allow assessment of the value of fresh tissue in ex vivo analysis as a biomarker of response.

This trial has been designed to minimise the chance of haematoma following surgery or biopsy in the presence of an anticoagulant. The terminal elimination half-life of Rivaroxaban is 5 to 9 h in healthy subjects aged 20 to 45 years and 11 to 13 h in elderly patients aged 60 to 76 years [60]. Clinical guidelines recommend stopping the drug in routine surgery patients at ≥ 24 h before surgery for minor risk procedures [61]; therefore, a safe time period, in patients with normal renal function, was taken as 24 h. However, clinicians (surgeons or radiologists performing biopsies) are still advised of the randomisation so bleeding precautions can be maximised.

The trial was initially designed as a window-of-opportunity pre-surgery study; however, with changing clinical practice, increasingly ER-negative patients are having neoadjuvant chemotherapy. Therefore, an amendment to the study was submitted to allow inclusion of neoadjuvant patients, with the post-trial biopsy being an additional research biopsy. It is routine to undergo radiological clip placement prior to commencement of chemotherapy, to allow identification of the tumour bed in the presence of complete radiological response to NAC. Therefore, to minimise impact on trial patients, the post-trial biopsy is performed at the time of clip placement. This was assessed as acceptable by our patient and public collaborators.