Keywords
cancer-associated thrombosis - low-molecular-weight heparin - rivaroxaban - routine
clinical practice - venous thromboembolism
Introduction
Rates of venous thromboembolism (VTE), comprising deep vein thrombosis (DVT) and pulmonary
embolism (PE), are particularly high in patients with cancer, and thrombosis can be
an indicator of occult cancer.[1] Agents used to treat cancer, such as chemotherapeutic agents or antiangiogenic agents,
have also been shown to increase the risk of thrombosis.[2]
[3]
[4] In cancer-associated thrombosis, treatment with low-molecular-weight heparin (LMWH)
was superior to vitamin K antagonists (VKAs) for the prevention of recurrence.[5] Guidelines, therefore, recommend extended (no scheduled stop date) LMWH monotherapy
over VKA therapy in this setting.[6] However, many physicians still use a traditional regimen of heparin (LMWH) followed
by a VKA to treat VTE in cancer-related thrombosis, for reasons including the high
cost of LMWHs over the longer term and injection discomfort for patients.[7]
Direct oral anticoagulants (DOACs) are as effective as standard anticoagulation for
the initial treatment of VTE and prevention of VTE recurrence.[8]
[9]
[10]
[11]
[12] A meta-analysis of six studies and a systematic review study showed that the DOACs
are as safe and effective as standard treatment for patients with VTE and cancer.[13] A pooled analysis from the EINSTEIN phase III studies of 1,124 patients with active
cancer (n = 655) or a history of cancer (n = 469) also showed that rivaroxaban had similar efficacy to enoxaparin plus a VKA
for preventing VTE recurrence, and rates of major bleeding were lower with rivaroxaban
than with enoxaparin/VKA.[14] However, none of the completed studies has directly compared DOACs with LMWH in
this setting, and guidelines still advise LMWH over DOACs as the treatment of choice
for VTE in patients with cancer.
The noninterventional XALIA phase IV study demonstrated the safety and effectiveness
of the single-drug approach with rivaroxaban for the treatment of DVT in routine clinical
practice.[15] The findings from XALIA were consistent with those from EINSTEIN DVT and, as with
the EINSTEIN phase III studies, several patients with cancer were included in the
XALIA study population. This analysis describes the baseline demographics and clinical
characteristics, treatment patterns, and outcomes of the patients with known and newly
diagnosed cancer in XALIA.
Methods
The methods were as described in the XALIA primary article.[15]
Study Design and Participants
XALIA was a multicenter, international, prospective, noninterventional study of patients
with objectively confirmed DVT. Patients received rivaroxaban or standard anticoagulation
treatment (initial treatment with unfractionated heparin, LMWH, or fondaparinux, usually
overlapping with and followed by a VKA). Patients could be included if they were 18
years or older, with objectively confirmed DVT, and an indication to receive anticoagulation
treatment for 3 months or longer. After the approval of rivaroxaban in the PE indication,
the protocol was amended to allow the enrolment of patients with DVT and concomitant
PE (but not isolated PE).
Cancer Reporting and Classification
In XALIA, the investigators reported cancer by marking checkboxes on the case report
forms (CRFs). Patients with cancer were therefore defined as having either known cancer
or cancer that was newly diagnosed at study entry. Cancers were categorized as breast,
central nervous system, gastrointestinal, genitourinary tract, lung, hematological
system, melanoma, or other. The cancer type was also recorded using a checkbox system.
If patients had more than one type of cancer, multiple checkboxes could be marked.
Procedures
Treatment, dose, and duration were at the attending physician's discretion. For the
purpose of this substudy, we defined the following treatment cohorts: rivaroxaban
cohort (patients treated with rivaroxaban alone or who received heparin or fondaparinux
for ≤48 hours before switching to rivaroxaban, consistent with the approach in EINSTEIN
DVT);[9] early switchers cohort (patients treated with rivaroxaban who received heparin or
fondaparinux for >48 hours to 14 days or a VKA for 1–14 days before changing to rivaroxaban);
standard anticoagulation cohort (patients treated with heparin or LMWH or fondaparinux
overlapping with and followed by a VKA); LMWH-alone cohort; and miscellaneous cohort
(patients treated with other heparins, fondaparinux alone, or a VKA alone). The observation
period ended 12 months from the date of the final patient enrolment; therefore, each
patient was followed up for at least 12 months.
Outcomes
The primary outcomes were as in XALIA: major bleeding, recurrent VTE, and all-cause
mortality. Major bleeding was defined as overt bleeding associated with a fall in
hemoglobin of ≥20 g/L; a transfusion of two or more units of packed red blood cells
or whole blood; critical site bleeding (intracranial, intraspinal, intraocular, pericardial,
intra-articular, intramuscular with compartment syndrome, and retroperitoneal); or
fatal bleeding. Recurrent VTE was defined as the new onset of symptoms confirmed by
diagnostic testing; only symptomatic VTE was considered in this study. Death was classified
as being VTE-related, bleeding-related, or from other causes. Secondary outcomes were
major adverse cardiovascular events (cardiovascular death, stroke, myocardial infarction,
unstable angina, acute coronary syndromes); other symptomatic thromboembolic events
(Budd-Chiari syndrome, retinal-vein thrombosis, sinus-vein thrombosis, portal-vein
thrombosis, catheter-associated thrombosis, upper-limb thrombosis [if initial DVT
was not an upper-limb thrombosis]); and other adverse events (AEs).
Statistical Analysis
A descriptive analysis was conducted and the crude rates for the primary outcomes
of the five treatment cohorts of patients with cancer were recorded. Due to imbalances
in clinical characteristics of the treatment groups, this analysis was purely observational
and no direct comparisons were conducted. Treatment patterns by type of cancer and
type of cancer treatment were also described. All analyses considered only outcomes
occurring on treatment (i.e., events were assigned to the treatment [rivaroxaban or
standard anticoagulation] which was given initially after the index event; the stopping
of this first medication was considered as the stopping of study medication). The
rationale for considering only events occurring on treatment was to focus on events
occurring under the newly introduced drug; for example, if a patient switched to rivaroxaban
on day 10 due to a bleeding event occurring earlier (e.g., on day 5), we considered
that such an event should not be assigned to the early switcher group, as it is not
related to the newly introduced drug.
Role of the Funding Source
Bayer AG and Janssen Research and Development, the XALIA study funders, gathered,
maintained, and extracted data. The authors had responsibility for data interpretation
and writing the article. All authors had access to the raw data. The corresponding
author had final responsibility to submit for publication. The study is registered
at www.clinicaltrials.gov (NCT01619007).
Results
Baseline Demographics and Clinical Characteristics
The total enrolment for XALIA between June 26, 2012, and March 31, 2014, was 5,142
patients; 6 patients did not take study medication and were excluded from the analysis.
Of the remaining 5,136 patients, 587 (11.4%) had known or newly diagnosed cancer at
baseline.
Of the 587 patients with cancer, 146 patients (24.9%) received rivaroxaban, 30 (5.1%)
were early switchers, 141 (24.0%) received standard anticoagulation, 223 (38.0%) received
LMWH alone, and 47 (8.0%) were in the miscellaneous cohort. Of the 30 early switchers
with cancer, 27 (90.0%) were switchers from heparin or fondaparinux and 3 (10.0%)
were switchers from a VKA. The flow of patients through the study is shown in [Fig. 1]; baseline demographics and clinical characteristics for patients with cancer are
shown for the five treatment cohorts in [Table 1]. The median treatment duration for patients with cancer was 152 days (interquartile
range [IQR]: 86–278 days) with rivaroxaban versus 164 days (IQR: 86–269 days) with
LMWH ([Table 2]).
Table 1
Baseline demographics and clinical characteristics of patients with cancer at baseline
|
Characteristic
|
Rivaroxaban
(n = 146)
|
Early switchers
(n = 30)
|
Standard anticoagulation
(n = 141)
|
LMWH
(n = 223)
|
Miscellaneous[a]
(n = 47)
|
|
Age, y, mean (SD)
|
69.3 (11.9)
|
70.4 (10.6)
|
70.4 (10.7)
|
68.0 (12.7)
|
70.0 (12.7)
|
|
Age category
|
|
< 60 y
|
29 (19.9)
|
4 (13.3)
|
22 (15.6)
|
52 (23.3)
|
6 (12.8)
|
|
≥ 60 y
|
117 (80.1)
|
26 (86.7)
|
119 (84.4)
|
171 (76.7)
|
41 (87.2)
|
|
Male
|
76 (52.1)
|
15 (50.0)
|
73 (51.8)
|
105 (47.1)
|
25 (53.2)
|
|
Weight
|
|
< 50 kg
|
5 (3.4)
|
0 (0.0)
|
1 (0.7)
|
12 (5.4)
|
2 (4.3)
|
|
≥ 50–70 kg
|
29 (19.9)
|
8 (26.7)
|
42 (29.8)
|
77 (34.5)
|
12 (25.5)
|
|
> 70– < 90 kg
|
55 (37.7)
|
12 (40.0)
|
53 (37.6)
|
71 (31.8)
|
12 (25.5)
|
|
≥ 90 kg
|
25 (17.1)
|
6 (20.0)
|
28 (19.9)
|
29 (13.0)
|
9 (19.1)
|
|
Missing
|
32 (21.9)
|
4 (13.3)
|
17 (12.1)
|
34 (15.2)
|
12 (25.5)
|
|
First available CrCl
|
|
< 30 mL/min
|
3 (2.1)
|
2 (6.7)
|
10 (7.1)
|
6 (2.7)
|
1 (2.1)
|
|
30– < 50 mL/min
|
14 (9.6)
|
5 (16.7)
|
13 (9.2)
|
15 (6.7)
|
6 (12.8)
|
|
50– < 80 mL/min
|
33 (22.6)
|
7 (23.3)
|
42 (29.8)
|
54 (24.2)
|
10 (21.3)
|
|
≥ 80 mL/min
|
44 (30.1)
|
9 (30.0)
|
36 (25.5)
|
67 (30.0)
|
14 (29.8)
|
|
Missing
|
52 (35.6)
|
7 (23.3)
|
40 (28.4)
|
81 (36.3)
|
16 (34.0)
|
|
Index diagnosis
|
|
DVT only
|
135 (92.5)
|
21 (70.0)
|
121 (85.8)
|
197 (88.3)
|
37 (78.7)
|
|
DVT with PE
|
11 (7.5)
|
9 (30.0)
|
20 (14.2)
|
26 (11.7)
|
10 (21.3)
|
|
Previous VTE
|
41 (28.1)
|
10 (33.3)
|
38 (27.0)
|
26 (11.7)
|
12 (25.5)
|
|
Known thrombophilic condition
|
5 (3.4)
|
0 (0.0)
|
8 (5.7)
|
6 (2.7)
|
1 (2.1)
|
|
Previous major bleeding episode
|
4 (2.7)
|
5 (16.7)
|
9 (6.4)
|
14 (6.3)
|
2 (4.3)
|
Abbreviations: CrCl, creatinine clearance; DVT, deep vein thrombosis; LMWH, low-molecular-weight
heparin; PE, pulmonary embolism; SD, standard deviation; VKA, vitamin K antagonist;
VTE, venous thromboembolism.
Note: Data are n (%) unless stated otherwise.
a Miscellaneous cohort included patients treated with other heparins, fondaparinux
alone, or a VKA alone.
Table 2
Treatment duration in patients with cancer
|
Treatment duration (d)
|
Rivaroxaban (n = 146)
|
Early switchers (n = 30)
|
Standard anticoagulation (n = 141)
|
LMWH (n = 223)
|
Miscellaneous[a] (n = 47)
|
|
Median
|
152
|
196
|
214
|
164
|
141
|
|
IQR
|
86–278
|
94–358
|
112–431
|
86–269
|
64–365
|
|
Mean (SD)
|
190.6 (161.2)
|
237.0 (195.1)
|
291.1 (214.2)
|
191.0 (145.6)
|
211.1 (192.9)
|
|
Min–max
|
1–749
|
1–859
|
1–898
|
2–751
|
1–714
|
Abbreviations: IQR, interquartile range; LMWH, low-molecular-weight heparin; SD, standard
deviation; VKA, vitamin K antagonist.
Note: Data are days unless stated otherwise.
a Miscellaneous cohort included patients treated with other heparins, fondaparinux
alone, or a VKA alone.
Fig. 1 Flow of patients through the study. A total of 5,142 patients were enrolled in XALIA;
6 patients were excluded as they did not take study medication. Of the remaining 5,136
patients, 587 (11.4%) had known or newly diagnosed cancer at baseline. Patients with
cancer were administered various anticoagulant treatment regimens, with the LMWH-treated
patients comprising the largest of the five study cohorts. *Early switchers were those
patients treated with rivaroxaban who received heparin or fondaparinux for >48 hours
to 14 days or a VKA for 1–14 days before changing to rivaroxaban. †Miscellaneous cohort included patients treated with other heparins, fondaparinux alone,
or a VKA alone. LMWH, low-molecular-weight heparin; VKA, vitamin K antagonist.
Reasons for Initial VTE Therapy Choice
Reasons for the choice of initial VTE treatment among patients with cancer are shown
in [Table 3]; more than one reason could be cited for this choice. In the rivaroxaban cohort,
the living conditions and age of the patient were the most frequently specified reasons;
guidelines, comorbidities, and drug availability were also commonly selected. For
the early switcher patients, medical/hospital guidelines were the most common reason,
and for the other three cohorts, the most frequently cited reason was comorbidities.
Table 3
Reason for choice of initial VTE treatment in patients with cancer[a]
|
Reason[a]
|
Rivaroxaban (n = 146)
|
Early switchers (n = 30)
|
Standard anticoagulation (n = 141)
|
LMWH (n = 223)
|
Miscellaneous[b] (n = 47)
|
|
Availability of drug
|
35 (24.0)
|
1 (3.3)
|
11 (7.8)
|
9 (4.0)
|
3 (6.4)
|
|
Comorbidities
|
38 (26.0)
|
9 (30.0)
|
70 (49.6)
|
170 (76.2)
|
25 (53.2)
|
|
Distance to treating physician
|
17 (11.6)
|
1 (3.3)
|
3 (2.1)
|
5 (2.2)
|
2 (4.3)
|
|
Medical or hospital guidelines
|
39 (26.7)
|
15 (50.0)
|
52 (36.9)
|
61 (27.4)
|
10 (21.3)
|
|
Patient's age
|
50 (34.2)
|
8 (26.7)
|
26 (18.4)
|
25 (11.2)
|
11 (23.4)
|
|
Patient's living condition
|
53 (36.3)
|
5 (16.7)
|
17 (12.1)
|
20 (9.0)
|
9 (19.1)
|
|
Price of drug
|
5 (3.4)
|
1 (3.3)
|
9 (6.4)
|
4 (1.8)
|
7 (14.9)
|
|
Type of health insurance
|
1 (0.7)
|
1 (3.3)
|
2 (1.4)
|
2 (0.9)
|
1 (2.1)
|
|
Other
|
20 (13.7)
|
4 (13.3)
|
13 (9.2)
|
18 (8.1)
|
10 (21.3)
|
Abbreviations: LMWH, low-molecular-weight heparin; VKA, vitamin K antagonist; VTE,
venous thromboembolism.
Note: Data are n (%) unless stated otherwise.
a More than one reason could be selected by the physician for each patient.
b Miscellaneous cohort included patients treated with other heparins, fondaparinux
alone, or a VKA alone.
Type of Cancer and Cancer Treatment Patterns
The most common type of cancer in the rivaroxaban, early switcher, standard anticoagulation,
and miscellaneous cohorts was genitourinary cancer (38/146 [26.0%] in the rivaroxaban
group, 12/30 [40.0%] in the early switchers, 53/141 [37.6%] in the standard anticoagulation
cohort, and 15/47 [31.9%] in the miscellaneous cohort; [Table 4]). The most common type of cancer in the LMWH group was gastrointestinal (57/223
[25.6%]). [Table 5] shows the proportion of patients who had received, or were currently receiving,
chemotherapy or hormone treatment, with the LMWH group having the highest proportion
of patients receiving ongoing treatment for cancer (99/223 [44.4%]).
Table 4
Known or newly diagnosed cancer at baseline
|
Cancer type
|
Rivaroxaban
(n = 146)
|
Early switchers
(n = 30)
|
Standard anticoagulation
(n = 141)
|
LMWH
(n = 223)
|
Miscellaneous[a]
(n = 47)
|
|
Breast
|
32 (21.9)
|
10 (33.3)
|
25 (17.7)
|
34 (15.2)
|
13 (27.7)
|
|
CNS
|
5 (3.4)
|
0 (0.0)
|
4 (2.8)
|
6 (2.7)
|
1 (2.1)
|
|
Gastrointestinal
|
20 (13.7)
|
4 (13.3)
|
19 (13.5)
|
65 (29.1)
|
8 (17.0)
|
|
Genitourinary
|
38 (26.0)
|
12 (40.0)
|
53 (37.6)
|
57 (25.6)
|
15 (31.9)
|
|
Hematological
|
12 (8.2)
|
0 (0.0)
|
13 (9.2)
|
22 (9.9)
|
3 (6.4)
|
|
Lung
|
5 (3.4)
|
1 (3.3)
|
10 (7.1)
|
23 (10.3)
|
5 (10.6)
|
|
Musculoskeletal
|
3 (2.1)
|
0 (0.0)
|
1 (0.7)
|
3 (1.3)
|
0 (0.0)
|
|
Melanoma
|
6 (4.1)
|
0 (0.0)
|
6 (4.3)
|
6 (2.7)
|
0 (0.0)
|
|
Other
|
33 (22.6)
|
4 (13.3)
|
20 (14.2)
|
21 (9.4)
|
7 (14.9)
|
Abbreviations: CNS, central nervous system; LMWH, low-molecular-weight heparin.
Note: Data are n (%) unless stated otherwise.
a Miscellaneous cohort included patients treated with other heparins, fondaparinux
alone, or a vitamin K antagonist alone.
Table 5
Chemotherapy and hormone therapy in patients with any cancer
|
Chemotherapy or hormone therapy in patients with any cancer
|
Rivaroxaban
(n = 146)
|
Early switchers
(n = 30)
|
Standard anticoagulation
(n = 141)
|
LMWH
(n = 223)
|
Miscellaneous[a]
(n = 47)
|
|
Previous therapy
|
33 (22.6)
|
11 (36.7)
|
38 (27.0)
|
44 (19.7)
|
14 (29.8)
|
|
< 6 mo before acute DVT
|
9 (6.2)
|
3 (10.0)
|
14 (9.9)
|
29 (13.0)
|
7 (14.9)
|
|
≥ 6 mo before acute DVT
|
24 (16.4)
|
8 (26.7)
|
24 (17.0)
|
15 (6.7)
|
7 (14.9)
|
|
Ongoing
|
21 (14.4)
|
3 (10.0)
|
25 (17.7)
|
99 (44.4)
|
16 (34.0)
|
Abbreviations: DVT, deep vein thrombosis; LMWH, low-molecular-weight heparin.
Note: Data are n (%) unless stated otherwise.
a Miscellaneous cohort included patients treated with other heparins, fondaparinux
alone, or a vitamin K antagonist alone.
Primary Outcomes
Results for the primary outcomes in patients with cancer are shown in [Fig. 2]. The rates of major bleeding were 1.4% (2/146) in the rivaroxaban group, 5.0% (7/141)
with standard anticoagulation, 3.6% (8/223) with LMWH, and 4.3% (2/47) in the miscellaneous
cohort; no major bleeding events occurred in the early switchers with cancer. The
rates of recurrent VTE were 3.4% (5/146) in the rivaroxaban group, 3.3% (1/30) in
the early switchers, 4.3% (6/141) in the standard anticoagulation group, 4.5% (10/223)
in the LMWH cohort, and 4.3% (2/47) in the miscellaneous cohort. All-cause mortality
occurred in 4.8% (7/146) of patients in the rivaroxaban group, 4.3% (6/141) of patients
in the standard anticoagulation group, 24.7% (55/223) of patients in the LMWH group,
and 14.9% (7/47) of patients in the miscellaneous cohort; there were no deaths in
the early switchers cohort. A detailed breakdown of the primary outcomes is shown
in [Table 6].
Table 6
Treatment-emergent clinical outcomes in patients with cancer
|
Outcome
|
Rivaroxaban
(n = 146)
|
Early switchers
(n = 30)
|
Standard anticoagulation
(n = 141)
|
LMWH
(n = 223)
|
Miscellaneous[a]
(n = 47)
|
|
Safety
|
|
Major bleeding episode (adjudicated)
|
|
Any
|
2 (1.4)
|
0 (0.0)
|
7 (5.0)
|
8 (3.6)
|
2 (4.3)
|
|
Fatal
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
1 (0.4)
|
0 (0.0)
|
|
Gastrointestinal
|
0 (0.0)
|
0 (0.0)
|
3 (2.1)
|
4 (1.8)
|
0 (0.0)
|
|
CNS (intracranial, subdural, subarachnoid, or cerebral)
|
2 (1.4)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
|
AEs
|
|
Any treatment-emergent AE
|
67 (45.9)
|
18 (60.0)
|
70 (49.6)
|
135 (60.5)
|
21 (44.7)
|
|
Any serious AE emerging during treatment
|
28 (19.2)
|
4 (13.3)
|
42 (29.8)
|
103 (46.2)
|
15 (31.9)
|
|
Any AE resulting in discontinuation of study drug
|
15 (10.3)
|
1 (3.3)
|
8 (5.7)
|
41 (18.4)
|
3 (6.4)
|
|
Any AE leading to or prolonging hospitalization
|
22 (15.1)
|
8 (26.7)
|
35 (24.8)
|
70 (31.4)
|
10 (21.3)
|
|
Effectiveness
|
|
Recurrent VTE
|
5 (3.4)
|
1 (3.3)
|
6 (4.3)
|
10 (4.5)
|
2 (4.3)
|
|
Type of recurrent VTE
|
|
Fatal PE
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
|
Death in which PE could not be ruled out
|
2 (1.4)
|
0 (0.0)
|
0 (0.0)
|
2 (0.9)
|
0 (0.0)
|
|
Nonfatal PE
|
2 (1.4)
|
0 (0.0)
|
2 (1.4)
|
4 (1.8)
|
0 (0.0)
|
|
Recurrent DVT plus PE
|
0 (0.0)
|
0 (0.0)
|
1 (0.7)
|
0 (0.0)
|
0 (0.0)
|
|
Recurrent DVT
|
0 (0.0)
|
1 (3.3)
|
3 (2.1)
|
5 (2.2)
|
2 (4.3)
|
|
Other
|
1 (0.7)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
|
Other
|
|
Major adverse cardiovascular events
|
1 (0.7)
|
0 (0.0)
|
1 (0.7)
|
2 (0.9)
|
0 (0.0)
|
|
Other thromboembolic events
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
|
All-cause mortality
|
7 (4.8)
|
0 (0.0)
|
6 (4.3)
|
55 (24.7)
|
7 (14.9)
|
|
Cause of death
|
|
VTE-related death
|
|
PE
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
|
PE not ruled out
|
2 (1.4)
|
0 (0.0)
|
0 (0.0)
|
1 (0.4)
|
0 (0.0)
|
|
Bleeding related
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
|
Cancer related
|
4 (2.7)
|
0 (0.0)
|
5 (3.5)
|
48 (21.5)
|
7 (14.9)
|
|
Cardiovascular disease
|
1 (0.7)
|
0 (0.0)
|
0 (0.0)
|
3 (1.3)
|
0 (0.0)
|
|
Other
|
0 (0.0)
|
0 (0.0)
|
1 (0.7)
|
3 (1.3)
|
0 (0.0)
|
Abbreviations: AE, adverse event; CNS, central nervous system; DVT, deep vein thrombosis;
LMWH, low-molecular-weight heparin; PE, pulmonary embolism; VTE, venous thromboembolism.
Note: Data are n (%) unless stated otherwise.
a Miscellaneous cohort included patients treated with other heparins, fondaparinux
alone, or a vitamin K antagonist alone.
Fig. 2 Primary outcomes in patients with cancer by treatment group. Outcomes are unadjusted
for imbalances in clinical characteristics at baseline, and cannot be directly compared
between treatment cohorts. *Miscellaneous cohort included patients treated with other
heparins, fondaparinux alone, or a VKA alone. LMWH, low-molecular-weight heparin;
VKA, vitamin K antagonist; VTE, venous thromboembolism.
Secondary Outcomes
Major Adverse Cardiovascular Events
Treatment-emergent major adverse cardiovascular events occurred in 0.7% (1/146) of
the rivaroxaban cohort, 0.7% (1/141) of the standard anticoagulation cohort, and 0.9%
(2/223) of the LMWH cohort; no events occurred in the early switcher or miscellaneous
cohorts.
Other Symptomatic Thromboembolic Events
There were no other symptomatic thromboembolic events in any of the treatment cohorts.
Other Adverse Events
In the rivaroxaban group, 45.9% (67/146) of patients had at least one treatment-emergent
AE. Of the early switchers, 60.0% (18/30) of patients had at least one AE. The proportion
in the standard anticoagulation group was 49.6% (70/141), 60.5% (135/223) in the LMWH
group, and 44.7% (21/47) in the miscellaneous group.
Discussion
As cancer therapies evolve and patients with tumors live longer, any additional information
regarding the treatment of cancer-associated diseases such as thrombosis in this clinically
challenging group of patients is a valuable addition to real-world practice and improving
patient outcomes. In XALIA, patients with known or newly diagnosed cancer at baseline
more frequently received parenteral treatment than rivaroxaban therapy. Some patients
with cancer (early switchers) were switched to rivaroxaban after an initial period
of treatment with standard anticoagulation. The largest of the five anticoagulation
treatment cohorts was the LMWH group, in accordance with current guidelines for cancer-associated
thrombosis.[6] These results suggested that, as expected, physicians were generally cautious in
prescribing a newer therapy to patients with cancer, in particular when chemotherapy
or hormone therapy was still ongoing. However, they were more willing to treat patients
with rivaroxaban or to switch some of the patients to rivaroxaban after initial treatment
with parenteral agents when chemotherapy was ended less than 6 months before the acute
DVT. In addition, many patients treated with standard anticoagulation received VKAs,
despite the guideline recommendations to treat cancer-associated thrombosis with LMWH;
the prescription of DOACs is advised as an alternative to VKAs.[6] Commonly cited reasons for VTE therapy choices included medical or hospital guidelines,
age of patient, availability of therapy, and the presence of comorbidities.
Patients with gastrointestinal cancer most often received LMWH; however, for other
sites, other treatments were more common. In the rivaroxaban group, the most common
sites of cancer were genitourinary or breast, and the proportion of patients receiving
rivaroxaban for these cancer types was higher than with LMWH. The LMWH group had the
highest proportion of patients receiving ongoing hormone therapy or chemotherapy during
the study—this may reflect the stage of the disease and may have contributed to the
higher rates of recurrent VTE and mortality observed in this group, because some cancer
treatments are associated with increased risks of thromboembolic events.[16]
[17]
[18]
For the primary outcomes of major bleeding, recurrent VTE, and all-cause mortality,
the rates were lower in patients treated with rivaroxaban than with standard anticoagulation,
LMWH, and miscellaneous cohorts (with the exception of all-cause mortality in the
standard anticoagulation cohort, which was lower than in the rivaroxaban cohort).
However, the study population was very heterogeneous and these comparisons were unadjusted,
meaning that the results were impacted by the baseline clinical characteristics of
the treatment group and therefore cannot be compared directly. Furthermore, the heterogeneity
of the study population could potentially have impacted treatment selection. Better-prognosis
patients may have been more likely to be administered rivaroxaban, increasing the
likelihood of lower outcome rates for this cohort. All-cause mortality was highest
by a substantial margin in the patients treated with LMWH, with death occurring in
a quarter of all patients; the most common cause of death in the latter group was
cancer (87.3% [48/55] of all deaths in the group), suggesting that patients with more
advanced-stage cancer were administered LMWH. VKA use (standard anticoagulation group)
was associated with the highest rate of major bleeding of the five cohorts, with similar
rates of VTE recurrence to patients receiving LMWH only. This is of interest, because
cancer-associated thrombosis is often treated with VKAs. Rates for all three primary
outcomes were higher in the rivaroxaban cohort for patients with cancer than in the
respective groups in the primary XALIA analysis (1.4 vs. 0.7% for major bleeding,
3.4 vs. 1.4% for recurrent VTE, and 4.8 vs. 0.5% for mortality); similarly, patients
in this substudy who received standard anticoagulation, LMWH alone, or miscellaneous
treatment had higher rates for all three primary outcomes versus the standard anticoagulation
treatment group in the main XALIA population.[15] This was similar to the findings of a substudy of patients with cancer versus those
without cancer from the EINSTEIN DVT and EINSTEIN PE studies,[14] and in other earlier studies,[19]
[20] where the presence of cancer was also associated with higher incidence rates for
the primary outcomes. These findings highlight the importance of subanalyses of patients
with and without cancer. In the early switchers cohort, primary outcome rates were
either very low or no events occurred, although the sample size for this group (30
patients) was very small. For the secondary outcomes, rates of major adverse cardiovascular
events were very low in all cohorts and no events occurred in the early switchers
or the miscellaneous cohort; no other symptomatic thromboembolic events occurred in
any of the treatment groups.
In terms of implications for clinical practice, the current study provides information
on how physicians consider treatment selection, and how outcomes in cancer-associated
thrombosis with a particular therapy are affected by the prognosis of the patient
population. Furthermore, the low rates for the primary outcomes with rivaroxaban in
patients with cancer-associated thrombosis in XALIA, and the fact that the observations
were consistent with the findings from the analysis of patients with cancer and VTE
in the EINSTEIN studies, should be reassuring to clinicians.
There were some limitations with this analysis: the small number of patients with
cancer, particularly in the early switcher group, means that the results must be interpreted
with caution (e.g., in the EINSTEIN studies, 651/8,246 [7.9%] patients had active
cancer in the safety analyses).[14] The imbalances in baseline clinical characteristics between treatment cohorts coupled
to the small sample size also meant that direct comparisons between treatment cohorts
were not appropriate; hence, our data are presented as descriptive only. In addition,
XALIA was an open-label study, which raised the possibility of bias in investigator
reporting of events; however, this was alleviated by the use of objective diagnostic
methods and by the adjudication of all events by the Adjudication Committee, which
was blinded to the treatment choice. We collected information on the presence of cancer,
but detailed information on activity of cancer is not available; however, based on
data for cancer treatment ([Table 5]), we can assume that the majority of patients had active cancer. Additionally, information
relating to cancer stage was not collected, which may have impacted the outcomes and
therapeutic decisions depending on the distribution of early- and late-stage cancers
between the treatment groups. An implicit criterion for inclusion in the study was
also life expectancy beyond around 3 months, as patients were required to have an
indication for at least 12 weeks of anticoagulant treatment. Therefore, this also
impacted mortality outcomes as patients with shorter life expectancies were excluded.
This was consistent with what was done in the EINSTEIN DVT study. Finally, values
for approximately one-third of creatinine clearance measurements were missing (in
cases where measurements were not recorded on the CRF, concerted efforts were made
to determine whether the measurements had not been taken, or they had been taken but
not recorded on the CRF by the investigator). Therefore, safety data must be assessed
cautiously, as both LMWH and rivaroxaban are renally excreted; however, as we have
discussed previously, this is not unexpected in a noninterventional study.[21]
In summary, this substudy of the noninterventional XALIA study provides evidence that
physicians treat cancer-associated thrombosis with different anticoagulant regimens.
The most commonly prescribed treatment regimen was LMWH. The high mortality in the
LMWH cohort suggests that physicians administer LMWH to patients with the most advanced
cancer. Rivaroxaban and standard anticoagulation are used in patients with better
prognosis and the early switchers comprise an intermediate prognosis group of patients.
Switching to rivaroxaban after parenteral anticoagulation for several days can also
be seen as a careful treatment approach to patient populations who were excluded from
the pivotal randomized trials. Finally, it would be of interest in the future to conduct
studies in populations with well-matched baseline characteristics to enable direct
comparisons of rivaroxaban to LMWH in patients with cancer-associated thrombosis.
