Patient-reported outcomes associated with changing to rivaroxaban for the treatment of cancer-associated venous thromboembolism – The COSIMO study

Abstract

Venous thromboembolism (VTE) is a frequent complication of cancer usually managed by anticoagulation therapy [[1]Cohen A.T. Katholing A. Rietbrock S. Bamber L. Martinez C. Epidemiology of first and recurrent venous thromboembolism in patients with active cancer. A population-based cohort study.Thromb. Haemost. 2017; 117: 57-65Crossref PubMed Scopus (106) Google Scholar]. Long-term daily doses of anticoagulation therapy may add to the emotional stress and treatment burden experienced by patients with cancer [[2]Noble S. Prout H. Nelson A. Patients&apos; Experiences of LIving with CANcer-associated thrombosis: the PELICAN study.Patient Prefer. Adherence. 2015; 9: 337-345Crossref PubMed Scopus (56) Google Scholar]. Observational data demonstrate that long-term persistence with LMWH therapy is low, suggesting general dissatisfaction among patients possibly because of the inconveniences associated with this therapy [[3]Ageno W. Samperiz A. Caballero R. Dentali F. Di Micco P. Prandoni P. et al.Duration of anticoagulation after venous thromboembolism in real world clinical practice.Thromb. Res. 2015; 135: 666-672Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar]. Changing from traditional anticoagulants (low-molecular-weight heparin, fondaparinux, or a vitamin K antagonist) to a guideline recommended direct oral anticoagulant (DOAC) may lessen the treatment burden associated with cancer-associated thrombosis (CAT) [[4]Khorana A.A. Noble S. Lee A.Y.Y. Soff G. Meyer G. O'Connell C. et al.Role of direct oral anticoagulants in the treatment of cancer-associated venous thromboembolism: guidance from the SSC of the ISTH.J. Thromb. Haemost. 2018; 16: 1891-1894Crossref PubMed Scopus (194) Google Scholar]. The convenience of a long-term therapy has a positive impact on treatment satisfaction, which can ultimately improve long-term compliance to, or persistence with, therapy and clinical outcomes [[5]Amin A. Marrs J.C. Direct oral anticoagulants for the management of thromboembolic disorders: the importance of adherence and persistence in achieving beneficial outcomes.Clin. Appl. Thromb. Hemost. 2016; 22: 605-616Crossref PubMed Scopus (17) Google Scholar]. The COSIMO study (NCT02742623) was designed to assess patient-reported treatment satisfaction following a planned change from traditional therapy to rivaroxaban for the treatment of CAT. The rationale and design of this study have been reported previously [[6]Cohen A.T. Maraveyas A. Beyer-Westendorf J. Lee A.Y.Y. Mantovani L.G. Bach M. et al.COSIMO - patients with active cancer changing to rivaroxaban for the treatment and prevention of recurrent venous thromboembolism: a non-interventional study.Thromb. J. 2018; 16: 21Crossref PubMed Scopus (14) Google Scholar]. It was a prospective, non-interventional, single-arm cohort study enrolling patients from 55 sites in Australia, Canada, and Europe. All patients provided written informed consent. The study was performed in accordance with the provisions of the Declaration of Helsinki and local regulations. Inclusion criteria were adult patients with active cancer other than fully treated basal-cell or squamous-cell carcinoma of the skin (with active cancer defined as the diagnosis or treatment of cancer within the previous 6 months or recurrent or metastatic cancer) and an Eastern Cooperative Oncology Group (ECOG) score < 3. Exclusion criteria included contraindications to rivaroxaban and venous thromboembolic events that occurred while patients were taking anticoagulant therapy. Eligible patients with acute venous thromboembolism who received traditional anticoagulant therapy for ≥4 weeks and were changing to rivaroxaban were included. The follow-up period was 6 months, with Anti-Clot Treatment Scale (ACTS) questionnaires completed at baseline and at approximately week 4 and months 3 and 6 for pairwise comparison to mean scores at baseline. The primary outcome was change in ACTS Burdens score between baseline and week 4. The effect size is calculated by dividing the mean difference (MD) by the standard deviation (SD) of difference (e.g. MD/SD = 3.8/6.7 = 0.56). The minimal clinically important difference (MCID) is calculated by multiplying the 0.3 (effect size) by the SD. An effect size of ≥0.3 (mean difference) was considered clinically significant. An effect size of 0.2 was considered a small difference, 0.5 a moderate difference, and 0.8 a large difference [7Bamber L. Wang M.Y. Prins M.H. Ciniglio C. Bauersachs R. Lensing A.W. et al.Patient-reported treatment satisfaction with oral rivaroxaban versus standard therapy in the treatment of acute symptomatic deep-vein thrombosis.Thromb. Haemost. 2013; 110: 732-741Crossref PubMed Scopus (72) Google Scholar, 8Prins M.H. Bamber L. Cano S.J. Wang M.Y. Erkens P. Bauersachs R. et al.Patient-reported treatment satisfaction with oral rivaroxaban versus standard therapy in the treatment of pulmonary embolism; results from the EINSTEIN PE trial.Thromb. Res. 2015; 135: 281-288Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar]. Details of the study questionnaires have been published previously [[6]Cohen A.T. Maraveyas A. Beyer-Westendorf J. Lee A.Y.Y. Mantovani L.G. Bach M. et al.COSIMO - patients with active cancer changing to rivaroxaban for the treatment and prevention of recurrent venous thromboembolism: a non-interventional study.Thromb. J. 2018; 16: 21Crossref PubMed Scopus (14) Google Scholar]. The ACTS questionnaire is a patient self-reporting instrument that measures the negative and positive aspects of anticoagulation therapy on separate subscales for ‘Burdens’ (across 13 questionnaire items including one global item) [[6]Cohen A.T. Maraveyas A. Beyer-Westendorf J. Lee A.Y.Y. Mantovani L.G. Bach M. et al.COSIMO - patients with active cancer changing to rivaroxaban for the treatment and prevention of recurrent venous thromboembolism: a non-interventional study.Thromb. J. 2018; 16: 21Crossref PubMed Scopus (14) Google Scholar]. Patients express their level of agreement with a statement on a 5-point Likert scale. Scores are reverse-coded during analysis; consequently, higher scores indicate greater patient treatment satisfaction. Observations continued for 6 months or until withdrawal of consent, death, or loss to follow-up. Patients who received at least one dose of rivaroxaban and who completed the ACTS questionnaire at the timepoint for analysis were included. The null hypothesis was no change in mean ACTS Burdens score between baseline and week 4 in pairwise analysis; a 5% significance level was used for hypothesis testing. The change in the mean ACTS Burdens score was analysed using the Wilcoxon signed-rank test. The assumption of normality was tested using the Shapiro–Wilk test at the 0.10 level of significance. For missing items, imputation to the mean was used where >50% of the questions (>6 items for ACTS Burdens) were completed. The study enrolled 509 patients; 505 patients received rivaroxaban and were included in the study population. Patient demographics, general clinical characteristics and cancer-related characteristics (Table 1) were assessed at the time of enrollment (baseline) and were similar between patients included in the study population and the ACTS week 4 analysis set. Overall, 117 (23.2%) patients in the study population discontinued the study prematurely: 59 (11.7%) died, 21 (4.2%) withdrew consent, 17 (3.4%) were lost to follow-up, and 20 (4.0%) had ‘other’ reasons for premature discontinuation. The majority of patients in the study population changed to rivaroxaban from LMWH therapy (n = 488, 96.6%); 8 (1.6%) patients changed from a VKA, and 9 (1.8%) patients from fondaparinux.Table 1Demographics, clinical and cancer-related characteristics at baseline for the safety and ACTS week 4 analysis sets.Characteristic, n (%)Study population(N = 505)ACTS week 4Analysis set (N = 381)Age, years, mean ± SD64.0 ± 11.764.2 ± 11.9Male gender225 (44.6)172 (45.1)Bodyweight, kg, mean ± SD76.7 ± 17.076.6 ± 16.6 <70162 (32.1)126 (33.1) ≥70–89189 (37.4)145 (38.1) ≥9097 (19.2)72 (18.9) Missing57 (11.3)38 (10.0)First available CrCl, mL/min <15–<304 (0.8)3 (0.8) ≥30–<5042 (8.3)31 (8.1) ≥50–<80148 (29.3)116 (30.4) ≥80234 (46.3)180 (47.2) Missing77 (15.2)51 (13.4)Index diagnosisaIndex diagnosis and known thrombophilia refer to the date of diagnosis of the index VTE. DVT only229 (45.3)180 (47.2)Symptomatic181 (35.8)145 (38.1)Incidental48 (9.5)35 (9.2) PE only188 (37.2)138 (36.2)Symptomatic116 (23.0)84 (22.0)Incidental72 (14.3)54 (14.2) PE with DVT49 (9.7)37 (9.7)Symptomatic34 (6.7)28 (7.3)Incidental15 (3.0)9 (2.4) Catheter-associated DVT38 (7.5)26 (6.8) Missing1 (0.2)0 (0.0)Previous VTE (≤5 years)44 (8.7)34 (8.9)Known thrombophiliaaIndex diagnosis and known thrombophilia refer to the date of diagnosis of the index VTE.6 (1.2)4 (1.0)Previous major bleeding episode11 (2.2)10 (2.6)Cancer category Solid tumour449 (88.9)322 (84.5)Breast84 (16.6)66 (17.3)CNS-glioblastoma11 (2.2)10 (2.6)Head and neck8 (1.6)3 (0.8)Lung59 (11.7)40 (10.5)Gastrointestinal131 (25.9)96 (25.2)Gynaecological80 (15.8)56 (14.7)Genitourinary58 (11.5)46 (12.1)Malignant melanoma3 (0.6)2 (0.5)Sarcoma5 (1.0)5 (1.3)Other10 (2.0)8 (2.1) Haematological malignancy56 (11.1)49 (12.9) Metastatic disease of solid tumours245 (48.5)177 (46.5)ECOG performance status 0162 (32.1)127 (33.3) 1276 (54.7)210 (55.1) 263 (12.5)42 (11.0)Missing4 (0.8)2 (0.5)Status of cancer response Complete remission47 (9.3)37 (9.7) Partial remission38 (7.5)29 (7.6) Stable disease146 (28.9)108 (28.3) Relapsed disease/progressive disease89 (17.6)66 (17.3) Not evaluable35 (6.9)27 (7.1) Not done150 (29.7)114 (29.9)Systemic therapy178 (35.2)141 (37.0) Chemotherapy150 (29.7)116 (30.4) Hormonal therapy18 (3.6)17 (4.5) Targeted therapy15 (3.0)12 (3.1) Immunotherapy15 (3.0)13 (3.4) Other6 (1.2)4 (1.0)Radiotherapy79 (15.6)69 (18.1)Abbreviations: ACTS, Anti-Clot Treatment Scale; CrCl, creatinine clearance; DVT, deep vein thrombosis; PE, pulmonary embolism; SD, standard deviation; VTE, venous thromboembolism; CNS, central nervous system; ECOG, Eastern Cooperative Oncology Group.Note: Characteristics were assessed at the time of enrolment (baseline) unless otherwise stated.a Index diagnosis and known thrombophilia refer to the date of diagnosis of the index VTE. Open table in a new tab Abbreviations: ACTS, Anti-Clot Treatment Scale; CrCl, creatinine clearance; DVT, deep vein thrombosis; PE, pulmonary embolism; SD, standard deviation; VTE, venous thromboembolism; CNS, central nervous system; ECOG, Eastern Cooperative Oncology Group. Note: Characteristics were assessed at the time of enrolment (baseline) unless otherwise stated. Overall, patients received a median of 100 days or 3.3 months (interquartile range of 47–181 days or 1.5–6.0 months) of anticoagulation therapy for their VTE before changing to rivaroxaban. The most common reasons for changing to rivaroxaban were patient-dependent reasons, which included a desire to cease parenteral administration (n = 136, 26.9%), improve QoL (n = 94, 18.6%), general patient preference (n = 76, 15.0%); for 174 (34.5%) patients, it was their physician's decision to change their therapy. Most patients were treated with rivaroxaban for at least 3 months (n = 457, 90.5%), and at the end of the observation period, 302 (59.8%) remained on rivaroxaban. The median duration of rivaroxaban treatment was 176 days (interquartile range 105–189 days). For the ACTS analyses, results from 423 (83.8%) patients were valid for inclusion in the ACTS over time analysis set, 381 (75.4%) in the week 4 analysis set, 341 (67.5%) in the month 3 analysis set, and 253 (50.1%) in the month 6 analysis set. Mean ACTS Burdens scores were significantly higher at week 4 compared with baseline (55.6 vs 51.8 out of a maximum score of 60, respectively; p < 0.0001; effect size 0.6) in the ACTS week 4 analysis set (N = 381 patients), signifying a clinically significant increase in treatment satisfaction. This result persisted at months 3 and 6 (Fig. 1). The minimal clinically important difference (MCID) was 2.0. (0.3 (min. sig. effect size) times 6.71 (SD)). 59.3% (226/381) of patients achieved a MCID by week 4. It is noteworthy that 47% (179/381) and 22% (84/381) of the patients reported ACTS baseline score of ≥58 and 60 respectively, suggesting that patient-reported treatment satisfaction increased at week 4 despite almost half of the patient cohort having a high satisfaction score at baseline. A sensitivity analysis on the ACTS Burden score at Week 4 was conducted to investigate the potential impact on the study outcome of patients who dropped out from the study earlier than Week 4 due to other reasons than death due to cancer. The mean (SD) ACTS Burden subscale was 51.7 (7.31) at baseline, which increased by 3.0 (8.96) to 54.7 (8.22) at Week 4. The median increased from 54.0 to 57.0 by a median of 2.6. The increase was statistically significant. Thus, the sensitivity analysis confirmed the primary outcome and shows that patients that dropped out had no significant impact on the primary outcome. Data describing patient satisfaction with DOACs for the treatment of CAT are limited. The primary outcome was met with a significant improvement in patient treatment satisfaction on the ACTS Burdens subscale observed at week 4 following a change from LMWH, fondaparinux, or a VKA to rivaroxaban. The significant improvement from baseline persisted at months 3 and 6. The almost immediate increase in treatment satisfaction observed at week 4 is strongly supportive of the treatment-dependent nature of the change in satisfaction. The improvement in ACTS Burdens score at week 4 (a mean difference of 3.9) was similar to previously reported differences in ACTS Burdens scores in head-to-head comparisons of rivaroxaban and VKA therapy for the treatment of VTE [7Bamber L. Wang M.Y. Prins M.H. Ciniglio C. Bauersachs R. Lensing A.W. et al.Patient-reported treatment satisfaction with oral rivaroxaban versus standard therapy in the treatment of acute symptomatic deep-vein thrombosis.Thromb. Haemost. 2013; 110: 732-741Crossref PubMed Scopus (72) Google Scholar, 8Prins M.H. Bamber L. Cano S.J. Wang M.Y. Erkens P. Bauersachs R. et al.Patient-reported treatment satisfaction with oral rivaroxaban versus standard therapy in the treatment of pulmonary embolism; results from the EINSTEIN PE trial.Thromb. Res. 2015; 135: 281-288Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar]. The effect size in COSIMO of 0.6 was considered to be a clinically significant change, while the mean differences in the EINSTEIN DVT and EINSTEIN PE trials were equivalent to moderate effect sizes of 0.4 and 0.5, respectively [7Bamber L. Wang M.Y. Prins M.H. Ciniglio C. Bauersachs R. Lensing A.W. et al.Patient-reported treatment satisfaction with oral rivaroxaban versus standard therapy in the treatment of acute symptomatic deep-vein thrombosis.Thromb. Haemost. 2013; 110: 732-741Crossref PubMed Scopus (72) Google Scholar, 8Prins M.H. Bamber L. Cano S.J. Wang M.Y. Erkens P. Bauersachs R. et al.Patient-reported treatment satisfaction with oral rivaroxaban versus standard therapy in the treatment of pulmonary embolism; results from the EINSTEIN PE trial.Thromb. Res. 2015; 135: 281-288Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar]. The COSIMO study utilised a non-interventional approach; therefore, patient-reported outcomes were representative of patients with CAT who are likely to be selected for rivaroxaban therapy in routine practice. Similar to other trials, approximately half of patients enrolled in the current study had metastases and ~90% had solid tumours [[9]Giustozzi M. Agnelli G. Del Toro-Cervera J. Klok F.A. Rosovsky R.P. Martin A.C. et al.Direct oral anticoagulants for the treatment of acute venous thromboembolism associated with cancer: a systematic review and meta-analysis.Thromb. Haemost. 2020; 120: 1128-1136Crossref PubMed Scopus (42) Google Scholar]. However, fewer patients received concomitant systemic anticancer therapy and/or radiotherapy (42.8% in the COSIMO study, compared with 72.4 and 69.5% of patients in the Hokusai-VTE-Cancer and SELECT-D trials, respectively) [[9]Giustozzi M. Agnelli G. Del Toro-Cervera J. Klok F.A. Rosovsky R.P. Martin A.C. et al.Direct oral anticoagulants for the treatment of acute venous thromboembolism associated with cancer: a systematic review and meta-analysis.Thromb. Haemost. 2020; 120: 1128-1136Crossref PubMed Scopus (42) Google Scholar]. Although a major limitation of this study is its single-arm study design, a non-randomised two-armed design, comparing outcomes in patients changing to rivaroxaban with those persisting with traditional anticoagulants, builds in inherent bias in addition to confounders that cannot be overcome or mitigated with statistical adjustment. Several factors in the study design may have potentially influenced the reported patient satisfaction score. Firstly, patients selected were likely to be dissatisfied with their current treatment regimen, which may have led to an overestimation of patient satisfaction after treatment change. However, our findings were consistent with a recent study assessing treatment satisfaction switching from vitamin K antagonists to DOACs [[10]Toorop M.M.A. van Rein N. Nierman M.C. Vermaas H.W. Huisman M.V. van der Meer F.J.M. et al.Switching from vitamin K antagonists to direct oral anticoagulants: Treatment satisfaction and patient concerns.J. Thromb. Haemost. 2020; 18: 1390-1397Crossref PubMed Scopus (2) Google Scholar]. Although baseline ACTS Burdens scores were high, a statistically and clinically significant improvement in ACTS Burdens score was continuously observed at all timepoints of measure. Secondly, the numbers of patients included in each timepoint analysis should be considered. There were 423 patients at baseline (100.0%) who had a completed and valid ACTS questionnaire and 381 patients (90.1%) at Week 4. Patients who discontinued prematurely, including those who died, might have experienced different treatment satisfaction to those who remained in the study. Regardless, according to the sensitivity analysis of the ACTS results, a noticeable improvement in treatment satisfaction was still observed at week 4 when baseline scores from patients who had discontinued early from the study including those who died for reasons other than cancer were included. General limitations of the study include patients being surveyed in writing and by telephone might bias the types of patients participating. Patient satisfaction is a subjective measurement and the possibility that some patients misinterpreted the questionnaires cannot be excluded. Of note, because 96.6% of patients changed from LMWH therapy, the primary driver for the improvement in patient satisfaction may be the change from an injectable to an oral agent. Finally, the geographical distribution of the study was limited to Europe, Canada, and Australia. The COSIMO study demonstrates that patients with CAT who changed their VTE treatment from LMWH, fondaparinux, or VKA therapy to rivaroxaban in everyday clinical practice experienced an improvement in treatment satisfaction, in reducing patient-reported anticoagulation burden. Improved treatment satisfaction following a change to rivaroxaban for the treatment of CAT has conceivable positive implications for long-term persistence with therapy and clinical outcomes. The COSIMO study was funded by Bayer AG and Janssen Pharmaceuticals . This work has been supported by the Italian Ministry of Health Ricerca Corrente – IRCCS MultiMedica . KF, KA, YD, SF, LB, and MB are employees of Bayer.

Topics & Concepts