Thrombosis • Research Paper

Home-treatment of deep vein thrombosis in patients

with cancer

Walter Ageno Background and Objectives. Outpatient treatment of deep vein thrombosis (DVT) has
Rebecca Grimwood become a common practice. However, in some centers cancer patients with DVT are
Sara Limbiati excluded from home treatment because they have a higher risk of both bleeding and recur-
Francesco Dentali rent DVT. We performed a retrospective review of clinical practice patterns to assess the
Luigi Steidl rate of cancer patients who were deemed eligible for outpatient treatment of their DVT.
Philip S. Wells
Design and Methods. The charts of patients from the Thrombosis Units at two tertiary care
institutions were reviewed. All patients with objectively documented DVT at our institutions
are treated through the Thrombosis Units. Patients are treated as outpatients unless they
require admission for other medical problems, are actively bleeding or have pain that
requires parenteral narcotics. Outpatient treatment was with low molecular weight heparin
(LMWH) followed by warfarin or with LMWH alone.
Results. Over a period of almost four years there were 321 patients with cancer, 167
(52.5%) of whom had metastatic disease. The most frequent sites of cancer were geni-
tourinary tract (21.2%), breast (20.5%), and gastrointestinal system (18.4%). Treatment
with LMWH and warfarin was prescribed to 67% and LMWH alone to 33%. One hundred
and ninety-seven patients (61.4%) were entirely treated at home. There were no differ-
ences between patients treated at home and hospitalized patients with regard to gender,
mean age, site of cancer, presence of metastases, and treatment. After 3 months, recur-
rent thromboembolism occurred in 6.1% of patients treated at home and in 4.8% of hos-
pitalized patients (p=0.64), and major bleeding in 1.0% and 4.8%, respectively (p=0.03).
One hundred and sixty patients died (49.8%), 100 (50.7%) in the home treatment group
and 60 (48.4%) of the hospitalized patients.
Interpretation and Conclusions. Home treatment of DVT in cancer patients is safe and fea-
sible in almost two-thirds of cases. Outpatient management of antithrombotic treatment
did not increase the rate of adverse events, even if the stage of the disease was
advanced.
Key words: venous thromboembolism, cancer, home treatment

Haematologica 2005; 90:220-224

©2005 Ferrata Storti Foundation
From the *Department of Clinical ancer is one of the most common cality of home treatment of DVT. Despite
Medicine, University of Insubria,
Varese (WA, SL, FD, LS), Italy,
and °University of Ottawa,
Department of Medicine and Ottawa
C causes of venous thromboem-
bolism (VTE). The prevalence of
known cancer ranges from 10 to 20% at
this evidence, rates of hospitalization of
patients with acute DVT vary among
countries and remain high in many cen-
Health Research Institute (RG, the time when VTE is diagnosed.1 ters. Social factors and the lack of a clear
PSW), Ottawa, Canada Treatment of VTE in cancer patients is consensus on concomitant clinical condi-
problematic because of a two- to three- tions that should mandate in-hospital
Correspondence:
Dr. Walter Ageno, U.O. Medicina I, fold higher risk of recurrence despite con- treatment are the main reasons for such
Ospedale di Circolo, viale Borri 57 ventional anticoagulant therapy and differences. Home treatment of patients
21100 Varese, Italy. because of an increased risk of bleeding with underlying malignancies has been
E-mail: walter.ageno@uninsubria.it complications.2-4 reported,7,8,10,11 but to our knowledge there
Home treatment of deep vein thrombo- are no studies that specifically report on
sis (DVT) is a common practice in many the safety and feasibility of the outpatient
European and North American centers. management of acute DVT in patients
Two large clinical trials5,6 clearly showed with cancer. Patients with cancer often
the efficacy and safety of low molecular have concomitant medical problems that
weight heparins (LMWH) in the outpa- complicate initial anticoagulation care and
tient setting, and a number of reports from thus suggest hospitalization for DVT
clinical practice7,8 or further comparisons treatment would be preferable. On the
between in-hospital and home treatment9 other hand, these patients may be inclined
have subsequently confirmed the practi- towards home-treatment because of a

| 220 | haematologica/the hematology journal | 2005; 90(2)

 Outpatient treatment of deep vein thrombosis

potential positive impact on their quality of life. tients. In both centers warfarin treatment was moni-
When the choice between home treatment and hos- tored by the institutional Anticoagulation Clinic with
pitalization was offered to cancer patients with acute INR monitoring according to individual needs but
DVT who were clinically eligible for outpatient man- with a maximum of four weeks between INR meas-
agement, 100% chose home treatment.12 This sug- urements. The follow–up lasted three months; (ii)
gests that provided home treatment is safe it should LMWH was administered for the entire three-month
be considered for all cancer patients. treatment period. LMWH was administered in a full
In this study we assessed the feasibility of home therapeutic dose for the first month and at 50% to
treatment of acute DVT in cancer patients and com- 75% of the full dose during the second and third
pared the outcomes in the hospitalized and home- months. No predefined criteria were used to select
treated patients. one of the two treatment modalities. At both institu-
tions, patients were encouraged to call the units if
symptoms of DVT or pulmonary embolism wors-
Design and Methods ened or developed, if bleeding was observed, if new
medications were started or if any invasive proce-
The charts of patients with objectively document- dures were planned. All patients underwent an
ed DVT and active cancer (ongoing or palliative) who assessment at the Thrombosis Unit after 3 months.
were referred to the Thrombosis Unit of the
Ospedale di Circolo of Varese, Varese, Italy from Assessments
February 2000 to June 2003 and to the Ottawa We collected data on the patients’ age, gender, site
Hospital General Campus Thrombosis Unit, Ottawa, of DVT, concomitant, objectively confirmed sympto-
Canada from January 1999 to December 2003 were matic pulmonary embolism, time between diagnosis
reviewed. Data on 22 patients referred to the of cancer and diagnosis of DVT, site of cancer, pres-
Thrombosis Unit of Varese in the years 2000 and ence of metastases and the specifics of cancer treat-
2001 were previously published as a part of a ment. For hospitalized patients, the reason for admis-
prospective cohort study.11 sion and the mean duration of stay were document-
All patients presenting with acute DVT, diagnosed ed. Details of the antithrombotic treatment were
by means of compression ultrasound or venography, obtained, rates of recurrent DVT or pulmonary
are routinely evaluated for the home treatment pro- embolism, major or minor bleeding, and mortality at
gram at both institutions. Usually, the following cri- 3 months were also collected. Recurrent DVT was
teria are applied for hospital admission: illness that defined as the detection of a new thrombus or the
independently requires hospitalization, high risk of extension of the previous thrombus documented by
bleeding or active bleeding, pain requiring parenteral compression ultrasound. For pulmonary embolism,
narcotics, likelihood of poor compliance, or refusal of high probability lung scans or filling defects on con-
home treatment. Patients selected for the home treat- trast spiral computed tomographic scans were diag-
ment program are provided with teaching on the nostic. An intermediate probability lung scan was
treatment and potential complications of venous considered diagnostic of acute pulmonary embolism
thromboembolism by either the attending physician if a new thrombosis was also found on compression
or a clinic nurse. Patients or care-givers are taught ultrasonography of the legs and the patient had
how to perform the injections and are given an expla- appropriate pulmonary symptoms. Bleeding was
nation of oral anticoagulant therapy, with particular defined as major if it was intracranial or retroperi-
emphasis on bleeding risks and the importance of toneal, or if it was overt and associated with either a
regular monitoring. The same education is provided decrease in hemoglobin levels of at least 2.0 g/dL or
to the hospitalized patients at the time of discharge. a need for the transfusion of at least 2 units of packed
Treatment was provided in two ways: (i) full dose red blood cells. Bleeding was defined as minor if it
subcutaneous, weight-adjusted, once or twice daily was overt but did not meet the criteria for major
LMWH (enoxaparin, dalteparin, or nadroparin) for a bleeding.
minimum of 5 days with concomitant warfarin, with The following characteristics were subsequently
LMWH discontinued when the international normal- compared between outpatients and hospitalized
ized ratio (INR) reached the therapeutic range (2.0 to patients: age, gender, site of DVT, concomitant
3.0) for two consecutive days. Warfarin was started symptomatic pulmonary embolism, site of cancer,
within 24 hours of diagnosis. INR monitoring and presence of known metastases, and concomitant
subsequent dosage adjustments were performed ongoing therapies for cancer. The rates of recurrent
daily for hospitalized patients, after 2 or 3 days and VTE, major and minor bleeding events and death at 3
then according to individual needs for the outpa- months were the outcome measures compared

haematologica/the hematology journal | 2005; 90(2) | 221 |

W. Ageno et al.

between outpatients and inpatients. The mean age of

Table 1. Baseline characteristics and cancer site and stage for out-
the patients and the mean time between diagnosis of patients and inpatients.
cancer and diagnosis of DVT were compared by t
tests, all other comparisons were performed by χ2
Outpatients Inpatients
tests. A p value lower than 0.05 was considered sta-
tistically significant. Number 197 124

Male gender, n(%) 74 (37.5%) 55 (44.3%)

Results Age (mean), years 60.0 61.1

Age (range), years 16-93 16-87

We identified 321 patients with cancer and an
objectively documented acute DVT. Their baseline Concomitant pulmonary 23 (11.7%) 27 (21.8%)
characteristics are summarized in Table 1, separately embolism, n(%)
for patients treated at home or in hospital. The mean Mean time from cancer 26.6 months 23.4 months
time between diagnosis of cancer and DVT was 25.4 to DVT diagnosis
months, the most common site of cancer was the
Site, n(%)
genitourinary tract, and 52.5% of patients had Genitourinary 39 (19.8%) 29 (23.4%)
known metastases at the time of DVT diagnosis. Breast 45 (22.8%) 21 (16.9%)
Cancer treatment was ongoing in 59.5% of patients, Gastrointestinal 39 (19.8%) 20 (16.1%)
Lung 31 (15.7%) 24 (19.3%)
most of whom were receiving chemotherapy. One Hematologic 20 (10.1%) 11 (8.9%)
hundred and ninety-seven patients were entirely Brain 9 (4.6%) 6 (4.8%)
treated at home (61.4%). For the 124 (38.6%) Other 14 (7.1%) 14 (11.3%)
patients who required hospitalization, the mean hos- Metastases 106 (53.8%) 61 (49.2%)
pital stay was 11.6 days. The most common reasons
cited for hospital admission were investigation of Ongoing chemotherapy* 103/194 (53.1%) 44/123 (35.8%)
recurrent cancer (n=49), concomitant medical disor- Ongoing radiotherapy* 33/194 (17%) 15/122 (12.3%)
der (n=24), concomitant pulmonary embolism
(n=13), concomitant DVT in other sites (n=9), illness Ongoing hormone therapy* 28/196 (14.3%) 16/123 (13%)
that independently required hospitalization (n=8), LMWH/Warfarin 131 (66.5%) 84 (67.7%)
and pain requiring parenteral narcotics (n=5).
Treatment with LMWH and warfarin was adminis- LMWH alone 66 (33.5%) 40 (32.3%)
tered to 215 patients (67%), LMWH alone was *Information on concomitant therapy was not available for all patients.
administered to the remaining 106 patients (33%).
Considering both the outpatients and hospitalized
patients, recurrent DVT or pulmonary embolism Table 2. Outcome events at 3 months.
occurred in 18 patients (5.6%), major bleeding in 5
patients (1.5%), and minor bleeding in 10 patients Outpatients Inpatients
(3.1%). One hundred and sixty patients (49.8%) died
during the 3-month follow-up. Comparing the out- Number 197 124
Recurrent VTE 12* (6.1%) 6°(4.8%)
patients and the hospitalized patients the rate of Major bleeding 2 (1.0%) 6 (4.8%)
recurrent venous thromboembolic events was similar Minor bleeding 4 (2.0%) 6 (4.8%)
(6.1% vs 4.8% respectively; p=0.64) but major bleed- Mortality 100 (50.7%) 60 (48.4%)
ing events occurred more commonly in hospitalized VTE: venous thromboembolism; *11 events were proximal DVT and 1 was
patients (1.0% vs 4.8%; p=0.03). The mortality rate pulmonary embolism; °5 events were proximal DVT and 1 was pulmonary
embolism.
was high in both groups but there was no statistical
difference between the groups (Table 2).
When we compared patients treated at home with
patients treated in hospital, we found no differences
according to gender and mean age. Patients with dis- uted in the two groups. The tumor type/origin and
tal DVT were more likely to be treated at home (35% the proportion of patients with metastatic disease
and 24.2%, respectively, p=0.04), whereas patients were similar between patients treated at home and
with concomitant pulmonary embolism were more patients treated in hospital. There was no difference
likely to be hospitalized (11.7% and 21.8%, respec- in the mean time interval between cancer diagnosis
tively, p=0.015). Patients with upper limb DVT or and DVT diagnosis in the two groups. Of interest,
with bilateral lower limb DVT were equally distrib- patients receiving chemotherapy at the time of DVT

| 222 | haematologica/the hematology journal | 2005; 90(2)

 Outpatient treatment of deep vein thrombosis

diagnosis were more likely to be treated at home strength of our findings and makes comparisons of
than to be admitted to hospital (p=0.0107). Finally, event rates less reliable. However, all patients were
there was no difference in antithrombotic treatment regularly followed-up at both institutions and the
between the 2 groups. pertinent information was routinely collected.
Although the study is retrospective the data were
collected prospectively. Moreover, the management
Discussion of these patients was similar in the two centers. The
mortality rate was high and we cannot confirm how
Our results support the feasibility of providing many deaths were due to unsuspected pulmonary
home treatment to most patients with active cancer embolism but the attending physicians did not
who present with acute DVT. In our experience, record pulmonary embolism as a suspected cause of
nearly two-thirds of patients with malignancy can death and the death rates were similar in both the
be managed entirely as outpatients regardless of hospitalized and outpatient therapy groups.
age, site of cancer, presence of known metastases, The results of our study support the initial home
and ongoing cancer treatment. The predominant treatment management of many cancer patients
reasons for providing treatment in hospital were the with acute DVT, and should promote this approach
need to investigate for recurrent cancer, or because in more institutions. The results of a recent random-
of concomitant medical problems usually related to ized controlled trial in which dalteparin was used
complications of the cancer. Recurrent venous for six months for the antithrombotic management
thromboembolic rates were similar in the two of DVT in cancer patients may further increase the
groups. Major hemorrhage was more frequent in feasibility of outpatient management by enabling
the hospitalized patients but this is to be expected elimination of the need for INR monitoring.13
given that active bleeding and a high risk of bleed- Furthermore, dalteparin was shown to be signifi-
ing were reasons for exclusion from home treat- cantly more effective than oral anticoagulants in
ment. The rate of patients with serious clinical con- reducing the risk of recurrent VTE without increas-
ditions was high in both groups, as shown by the ing the risk of bleeding. Oral anticoagulants are par-
nearly identical rate of patients with metastases and ticularly problematic in cancer patients because of
the high mortality rate. Despite this being a serious- interactions with several chemotherapies, because
ly ill group of patients, home treatment of DVT of poor venous access or poor clinical conditions
was at least as safe and effective as in-hospital treat- that can make laboratory monitoring difficult, and
ment. because of more cumbersome management in the
To our knowledge, this is the first study that has case of bleeding or invasive procedures. Many
specifically addressed the initial outpatient manage- patients are commonly hospitalized for strict moni-
ment of DVT in cancer patients. We believe that toring of the induction phase of oral anticoagulant
cancer patients can derive important benefits from treatment when receiving chemotherapy or due to
outpatient treatment because a new hospital admis- concomitant illness that makes ambulatory INR
sion can negatively affect their quality of life. monitoring unfeasible.
However, cancer is frequently considered a criterion In conclusion, we found that most patients with
for exclusion from home treatment of DVT and in active malignancy and acute DVT can be safely and
many centers these patients are routinely admitted effectively treated at home even in the presence of
to hospital. In a previous study,11 we observed that advanced disease. Home treatment of DVT is safe
43% of hospital admissions for acute DVT were and feasible and is likely to affect the quality of life
due to concomitant malignancy. This study con- of cancer patients positively. Further randomized
firms that despite a higher risk of recurrent venous controlled studies comparing in-hospital treatment
thromboembolic events or hemorrhagic events in and home treatment of DVT may be warranted.
cancer patients, outpatient management of DVT is
WA and PSW were the major contributors to this work and were
possible in the majority, even if life expectancy is primarily responsible for it, from conception to submitted manu-
short. Based on the results of our study, patients script. All authors qualified for authorship according to the World
with less extensive disease, in particular patients Association of Medical Editors (WAME) criteria, and have taken
specific responsibility for the following parts of the content: RG, SL,
without concomitant pulmonary embolism, and FD, and LS collection of clinical data; LS had also a role in design-
patients with ongoing chemotherapy were the best ing the study, interpreting the data and preparing the article.
candidates for the outpatient treatment. Dr. Wells is the recipient of a Canada Research Chair in
Thrombosis award.
Our study has limitations. The retrospective Manuscript received September 2, 2004. Accepted December 2,
design and lack of randomization weakens the 2004.

haematologica/the hematology journal | 2005; 90(2) | 223 |

W. Ageno et al.

Meer J, et al. Treatment of venous Med 1998;158:1809-12.

References thrombosis with intravenous unfrac- 9. Boccalon H, Elias A, Chale JJ, Cadene
tionated heparin administered in the A, Gabriel S. Clinical outcome and
1. Otten HMMB, Prins MH. Venous hospital as compared with subcuta- cost of hospital vs home treatment of
thromboembolism and occult malig- neous low molecular weight heparin proximal deep vein thrombosis with a
nancy. Thromb Res 2001;102:V187-94. administered at home. N Engl J Med low molecular weight heparin. Arch
2. Lee AYY. Treatment of venous throm- 1996;334:682-7. Intern Med 2000;160:1769-73.
boembolism in cancer patients. 6. Levine M, Gent M, Hirsh J, Leclerc J, 10. Prandoni P. Antithrombotic strategies
Thromb Res 2001;102:V195-208. Anderson D, Weitz J, et al. A compar- in patients with cancer. Thromb
3. Hutten BA, Prins MH, Gent M, Gins- ison of low molecular weight heparin Hamost 1997;78:141-4.
berg J, Tijssen JG, Buller HR. Incidence administered primarily at home with 11. Ageno W, Steidl L, Marchesi C,
of recurrent thromboembolic and unfractionated heparin administered Dentali F, Mera V, Squizzato A, et al.
bleeding complications among in the hospital for proximal deep vein Selecting patients for home treatment
patients with venous thromboem- thrombosis. N Engl J Med 1996; 334: of deep vein thrombosis: the problem
bolism in relation to both malignancy 677-81. of cancer. Haematologica 2002;87:286-
and achieved international normalized 7. Harrison L, McGinnis J, Crowther M, 91.
ratio: a retrospective analysis. J Clin Ginsberg J, Hirsh J. Assessment of out- 12. Vitale J, Ageno W. Home treatment of
Oncol 2000;18:3078-83. patient treatment of deep vein throm- cancer patients with deep venous
4. Prandoni P, Lensing AW, Piccioli A, bosis with low molecular weight thrombosis. Pathophysiol Haemost
Bernardi E, Simioni P, Girolami B, et al. heparin. Arch Intern Med 1998; 158: Thromb 2003;33 Suppl 1:100.
Recurrent venous thromboembolism 2001-3. 13. Lee AYY, Levine MN, Baker RI,
and bleeding complications during 8. Wells PS, Kovacs MJ, Bormanis J, Bowden C, Kakkar AK, Prins M, et al.
anticoagulant treatment in patients Forgie MA, Goudie D, Morrow B, et Low molecular weight heparin versus
with cancer and venous thrombosis. al. Expanding eligibility for outpatient a coumarin for the prevention of recur-
Blood 2002;100:3484-8. treatment of deep venous thrombosis rent venous thromboembolism in
5. Koopman MMW, Prandoni P, Piovella and pulmonary embolism with low patients with cancer. N Engl J Med
F, Ockelford PA, Brandjes DP, van der molecular weight heparin. Arch Intern 2003;349:146-53.

| 224 | haematologica/the hematology journal | 2005; 90(2)