Professional Documents
Culture Documents
Khorana AA - Thromboembolism in Hospitalized Neutropenic Cancer Patients
Khorana AA - Thromboembolism in Hospitalized Neutropenic Cancer Patients
Khorana AA - Thromboembolism in Hospitalized Neutropenic Cancer Patients
net/publication/7352045
CITATIONS READS
411 467
6 authors, including:
Gary H Lyman
Fred Hutch Cancer Center
1,010 PUBLICATIONS 71,940 CITATIONS
SEE PROFILE
All content following this page was uploaded by Gary H Lyman on 30 December 2015.
484
Downloaded from jco.ascopubs.org on December 27, 2015. For personal use only. No other uses without permission.
Copyright © 2006 American Society of Clinical Oncology. All rights reserved.
Thromboembolism in Cancer Patients
www.jco.org 485
Downloaded from jco.ascopubs.org on December 27, 2015. For personal use only. No other uses without permission.
Copyright © 2006 American Society of Clinical Oncology. All rights reserved.
Khorana et al
Table 1. Characteristics of the Study Population and Association With Venous and Arterial Thromboembolism by Univariate Analysis
Venous Thromboembolism Arterial Thromboembolism
No. of % of % With Odds % With Odds
Characteristic Patients Total Event 95% CI Ratioⴱ 95% CI Event 95% CI Ratio 95% CI
All patients 66,106 100 5.37 5.20 to 5.54 1.47 1.38 to 1.56
Age, years
⬍ 65 48,750 73.75 5.08 4.89 to 5.28 1.14 to 1.33 0.97 0.89 to 1.06 2.64 to 3.40
ⱖ 65 17,356 26.25 6.18 5.82 to 6.54 1.23† 2.86 2.61 to 3.11 3.0†
Sex
Female 34,639 52.40 5.46 5.22 to 5.70 0.90 to 1.03 1.32 1.29 to 1.44 1.10 to 1.41
Male 31,467 47.60 5.27 5.02 to 5.52 0.96‡ 1.63 1.49 to 1.77 1.25‡
Race/ethnicity
White 48,161 72.85 5.35 5.15 to 5.55 0.99 0.92 to 1.07 1.44 1.33 to 1.55 0.94 0.81 to 1.08
Black 7,438 11.25 5.54 5.02 to 6.06 1.04 0.93 to 1.15 1.75 1.45 to 2.05 1.23 1.02 to 1.48
Hispanic 3,075 4.65 5.11 4.35 to 5.88 0.95 0.80 to 1.12 1.07 0.71 to 1.44 0.72 0.51 to 1.02
Other/unknown 7,432 11.24 5.42 4.91 to 5.94 1.01 0.91 to 1.13 1.52 1.24 to 1.80 1.04 0.86 to 1.27
Site of cancer
Leukemia 14,600 22.09 4.39 4.06 to 4.72 0.77 0.70 to 0.84 1.91 1.69 to 2.13 1.43 1.25 to 1.65
Non-Hodgkin’s lymphoma 12,977 19.63 5.01 4.63 to 5.38 0.91 0.84 to 1.00 1.33 1.14 to 1.53 0.89 0.75 to 1.05
Breast 5,187 7.85 3.93 3.42 to 4.50 0.70 0.61 to 0.81 0.69 0.47 to 0.92 0.45 0.32 to 0.63
Lung 4,655 7.04 7.00 6.27 to 7.74 1.36 1.21 to 1.53 2.81 2.34 to 3.29 2.09 1.74 to 2.52
Myeloma 4,527 6.85 5.79 5.11 to 6.47 1.09 0.96 to 1.24 0.93 0.65 to 1.21 0.61 0.45 to 0.84
Hodgkin’s disease 2,042 3.09 3.87 3.03 to 4.71 0.70 0.56 to 0.88 0.54 0.22 to 0.86 0.36 0.20 to 0.65
Ovary 1,955 2.96 6.50 5.40 to 7.39 1.23 1.03 to 1.48 1.38 0.86 to 1.90 0.94 0.64 to 1.38
Head and neck 1,606 2.43 2.74 1.94 to 3.54 0.49 0.36 to 0.66 1.81 1.15 to 2.46 1.24 0.86 to 1.80
Sarcoma 1,597 2.42 4.82 3.77 to 5.87 0.89 0.71 to 1.12 0.31 0.04 to 0.59 0.21 0.09 to 0.50
Endometrium/cervical 1,060 1.60 8.96 7.24 to 10.68 1.76 1.42 to 2.17 1.42 0.70 to 2.13 0.96 0.58 to 1.61
Colon 756 1.14 6.75 4.96 to 8.53 1.28 0.96 to 1.70 2.65 1.50 to 3.79 1.84 1.18 to 2.89
Other abdominal 720 1.09 7.64 5.70 to 9.58 1.47 1.11 to 1.93 1.39 0.53 to 2.24 0.95 0.51 to 1.77
Brain 716 1.08 9.50 7.35 to 11.64 1.87 1.45 to 2.40 0.70 0.09 to 1.31 0.47 0.19 to 1.13
Esophagus 699 1.06 6.72 4.87 to 8.58 1.27 0.95 to 1.72 2.00 0.96 to 3.04 1.38 0.81 to 2.35
Rectum 645 0.98 6.98 5.13 to 9.22 1.33 0.98 to 1.80 1.86 0.82 to 2.90 1.28 0.72 to 2.27
Stomach 580 0.88 7.41 5.28 to 9.55 1.42 1.04 to 1.94 1.72 0.66 to 2.78 1.18 0.63 to 2.21
Testicular 491 0.74 6.52 4.33 to 8.70 1.23 0.86 to 1.76 0.81 0.02 to 1.61 0.55 0.21 to 1.47
Prostate 439 0.66 7.29 4.86 to 9.72 1.39 0.97 to 2.00 3.87 2.07 to 5.68 2.74 1.68 to 4.46
Pancreas 438 0.66 12.10 9.05 to 15.15 2.45 1.84 to 3.27 1.60 0.42 to 2.77 1.09 0.52 to 2.31
Bladder 288 0.44 6.60 3.73 to 9.46 1.25 0.78 to 1.99 2.78 0.88 to 4.68 1.93 0.95 to 3.90
Renal 212 0.32 7.55 3.99 to 11.10 1.44 0.87 to 2.40 0.47 0.00 to 1.39 0.32 0.05 to 2.27
Multiple cancers 906 1.37 4.97 3.55 to 6.38 0.92 0.68 to 1.24 2.43 1.43 to 3.43 1.69 1.10 to 2.59
Other cancers 9,010 13.63 5.98 5.49 to 6.47 1.14 1.04 to 1.26 1.02 0.81 to 1.23 0.66 0.53 to 0.82
Comorbidities
Infection 28,904 43.72 6.32 6.04 to 6.60 1.39 1.30 to 1.49 1.85 1.69 to 2.00 1.59 1.40 to 1.80
Pulmonary disease 13,100 19.82 8.05 7.59 to 8.52 1.77 1.65 to 1.91 3.18 2.88 to 3.48 3.11 2.73 to 3.53
Hypertension 12,463 18.85 5.89 5.48 to 6.30 1.13 1.04 to 1.23 2.27 2.01 to 2.53 1.79 1.56 to 2.06
Renal disease 7,549 11.42 7.91 7.30 to 8.52 1.62 1.48 to 1.77 3.30 2.90 to 3.70 2.74 2.36 to 3.17
Diabetes mellitus 5,122 7.75 5.62 4.99 to 6.25 1.05 0.93 to 1.19 3.05 2.58 to 3.52 2.32 1.95 to 2.76
Congestive heart failure 2,722 4.12 7.02 6.06 to 8.04 1.35 1.16 to 1.57 7.35 6.37 to 8.33 6.45 5.49 to 7.57
Hepatic disease 2,090 3.16 7.37 6.25 to 8.49 1.42 1.20 to 1.68 2.11 1.49 to 2.72 1.47 1.08 to 1.99
Obesity 518 0.78 8.11 5.76 to 10.46 1.56 1.14 to 2.15 1.74 0.61 to 2.86 1.19 0.61 to 2.31
Past history
Venous thromboembolism 878 1.33 3.87 2.60 to 5.15 0.71 0.50 to 1.00 1.25 0.52 to 1.99 0.85 0.47 to 1.55
Chronic anticoagulation 397 0.6 15.87 12.41 to 19.46 3.37 2.57 to 4.42 1.76 0.71 to 3.6 1.2 0.57 to 2.56
ⴱ
An odds ratios of ⬎ 1.0 for age signifies a greater likelihood of thromboembolism in neutropenic cancer patients age ⱖ 65 years, and, for sex, signifies a greater
likelihood of thromboembolism in male patients. For sites of cancer, an odds ratio of ⬎ 1.0 signifies a greater likelihood of thromboembolism relative to other
neutropenic cancer patients.
†For age ⱖ 65 years.
‡For male sex.
Any thromboembolism 4,434 6.71 6.52 to 6.9 838 5.69 5.32 to 6.06 5,272 7.98 7.77 to 8.18
Venous thromboembolism 3,550 5.37 5.2 to 5.54 705 4.79 4.44 to 5.13 4,255 6.44 6.25 to 6.62
Venous thrombosis 3,167 4.79 4.63 to 4.95 661 4.49 4.15 to 4.82 3,828 5.79 5.61 to 5.97
Pulmonary embolism 616 0.93 0.86 to 1.01 95 0.65 0.52 to 0.77 711 1.08 1 to 1.15
Arterial thromboembolism 970 1.47 1.38 to 1.56 165 1.12 0.95 to 1.29 1,135 1.72 1.62 to 1.82
Acute arterial embolism 141 0.21 0.18 to 0.25 30 0.20 0.13 to 0.28 171 0.26 0.22 to 0.3
Acute cerebrovascular disease 353 0.53 0.48 to 0.59 69 0.47 0.36 to 0.58 422 0.64 0.58 to 0.7
Acute coronary artery disease 506 0.77 0.7 to 0.83 71 0.48 0.37 to 0.59 577 0.87 0.8 to 0.94
greater mortality than patients without such a diagnosis (OR ⫽ 2.01; significance in this model were year of hospitalization (with mor-
95% CI, 1.83 to 2.22; P ⬍ .0001), and a similar trend was observed in tality declining over the duration of study), age ⱖ 65 years, race,
the subgroup of patients with information regarding metastatic dis- site of cancer, and comorbidities.
ease (Fig 2A). In-hospital mortality was greater in patients with venous
thromboembolism and either nonmetastatic disease (OR ⫽ 1.62; 95%
CI, 1.37 to 1.91; P ⬍ .0001) or metastatic disease (OR ⫽ 2.06; 95% CI, DISCUSSION
1.74 to 2.44; P ⬍ .0001). Similarly, in-hospital mortality was greater in
patients with arterial thromboembolism (OR ⫽ 5.05; 95% CI, 4.38 to We studied the occurrence of thromboembolism in hospitalized
5.79) and either nonmetastatic disease (OR ⫽ 3.6; 95% CI, 2.96 cancer patients receiving chemotherapy, as identified by the pres-
to 4.37; P ⬍ .0001) or metastatic disease (OR ⫽ 3.87; 95% CI, 2.93 to ence of neutropenia. The analysis shows that venous and arterial
5.13; P ⬍ .0001; Fig 2B). thromboembolic events are frequent complications of hospitaliza-
Multivariate Analysis tion in this population and that they both contribute to increased
The following clinical variables were found to be significantly in-hospital mortality. The proportion of patients with thrombo-
associated with venous thromboembolism using a multivariate embolism is high across all subgroups studied including patients
logistic regression analysis: age ⱖ 65 years; site of cancer, including with hematologic malignancies and with nonmetastatic disease,
brain, lung, stomach, pancreas, other abdominal, ovary, endome- who are commonly perceived to be at lower risk for thromboem-
trium, and cervix; arterial thromboembolism; and presence of bolism. Older age, particular sites of cancer, infection, and other
comorbidities, including infection, pulmonary and renal disease, comorbidities are associated with a higher risk of thromboembolic
and obesity. Their association with venous thromboembolism is events. Of particular concern is the increasing frequency of venous
described in Table 4. A similar multivariate analysis for mortality and arterial thromboembolism.
identified venous (OR ⫽ 1.5; 95% CI, 1.3 to 1.6) and arterial throm- The proportion of patients with venous thromboembolism re-
boembolism (OR ⫽ 2.7; 95% CI, 2.3 to 3.2) as significantly associ- ported in our study is greater than the proportion observed in a
ated with increased in-hospital mortality. Other variables of prospective study of acutely ill medical patients, which found an inci-
dence of 0.7% symptomatic venous thrombosis and 1% pulmonary
embolism in the placebo arm.9 The proportion is also much higher
than the 0.6% incidence observed in Medicare claims data from 1988
to 1990.6 The incidence of arterial thromboembolism in hospitalized
cancer patients has not been previously reported. Several factors may
have contributed to the high proportion of thromboembolism ob-
served. The study population comprised only patients receiving
chemotherapy because a discharge diagnosis of neutropenia was
required. Nearly 44% of patients experienced an infectious com-
plication during hospitalization. Both chemotherapy and infection
are risk factors for thrombosis.3,10-12 Hematopoietic growth fac-
tors, which are commonly used in hospitalized patients with febrile
neutropenia, may be thrombogenic, although a recent meta-
analysis was inconclusive.13 The true risk of thromboembolism
may have increased since the last report; indeed, we observed a
significant increase in its occurrence over the study duration. There
may be an increased awareness of the diagnosis of thromboembo-
Fig 1. Increase in the proportion of patients with venous and arterial thrombo- lism, leading to increased testing.14 Risk factors, such as patient age
embolism over time. Results are presented as annual proportions between 1995
and 2002. Significant trends for increase in occurrence were observed for both and obesity, increased over the same period, although length of hos-
venous and arterial thromboembolism (P ⬍ .0001 for trend). pitalization declined. Newer, more thrombogenic cancer treatment
www.jco.org 487
Downloaded from jco.ascopubs.org on December 27, 2015. For personal use only. No other uses without permission.
Copyright © 2006 American Society of Clinical Oncology. All rights reserved.
Khorana et al
regimens, increasing use of implanted venous access devices, and shown to be associated with an advanced stage and a three-fold lower
an increased awareness of thrombosis as a complication of hospi- survival at 1 year.26
talization may also have contributed.15-17 Although the discharge Our study had several limitations. It was retrospective and based
codes we used to identify venous thrombosis included superficial on discharge codes. However, accuracy of coding for pulmonary em-
thrombophlebitis, the number of patients with this diagnosis alone bolism and deep venous thrombosis has been previously shown to be
was small (n ⫽ 198, 0.2%). 92% and 84%, respectively.27 Neither the time to thromboembolic
The significance of cancer site suggests that intrinsic factors event nor the time from thromboembolism to mortality were avail-
unique to these particular tumors contribute to hypercoagulabil- able. The study included only inpatients and may have underesti-
ity. The high risk of venous thromboembolism in patients with mated the true proportion because venous thromboembolism is often
pancreas, brain, and gynecologic cancers is consistent with prior diagnosed and managed on an outpatient basis. Although neutrope-
observations.7,18-20 The association of arterial thromboembolism nia is usually a manifestation of chemotherapy administration, it can
with prostate, lung, and colon cancers and leukemia is a novel also occur without the use of chemotherapy, particularly in patients
finding and deserves further investigation. with myeloproliferative disorders. In this study, however, the propor-
The occurrence of thromboembolism has several clinical conse- tion of patients with thromboembolism was not substantially altered
quences related to patient morbidity, interruption of chemotherapy, when patients with myeloproliferative disorders were excluded (data
and cost of hospitalization.21 In our analysis, thromboembolism was not shown). We could not validate the diagnosis of neutropenia in our
also associated with increased in-hospital mortality. Arterial throm- study population, although it has been previously shown to be specific
boembolism and pulmonary embolism can be life threatening and in similar administrative datasets.28,29 Thromboprophylaxis can alter
could have contributed to the observed increase in mortality. How- the incidence of venous thromboembolism, but its use in our study
ever, venous thrombosis without embolism is not a life-threatening population was not known.30 According to a recent survey, however,
condition. Experimental models suggest that activation of the coagu- less than 5% of medical oncology patients are routinely administered
lation cascade confers a growth advantage on the tumor. In particular, thromboprophylaxis.31 Venous thromboembolism occurred in 5.4%
tissue factor, thrombin, and fibrin(ogen) can enhance tumor growth, of patients, and therefore, logistic regression analysis may be ham-
metastasis, and angiogenesis.22-25 The presence of venous thrombosis pered by the rare disease assumption. Information regarding stage of
could signify an aggressive tumor biology and, therefore, a worse disease, use of implanted intravascular access devices, type and timing
short-term prognosis. Cancer diagnosed at the same time as or within of chemotherapy regimens, and outside-hospital mortality was not
1 year of an episode of venous thromboembolism has previously been available. Documentation regarding other known risk factors for
9. Samama MM, Cohen AT, Darmon JY, et al: A phate, paclitaxel, and carboplatin in patients with
REFERENCES comparison of enoxaparin with placebo for the pre- castrate, metastatic prostate adenocarcinoma. Can-
vention of venous thromboembolism in acutely ill cer 98:1842-1848, 2003
1. Trousseau A, Bazire PV, Cormack JR: Lec- medical patients: Prophylaxis in medical patients 16. Zangari M, Barlogie B, Thertulien R, et al:
tures on Clinical Medicine. London, United King- with enoxaparin study group. N Engl J Med 341:793- Thalidomide and deep vein thrombosis in multiple
dom, R. Hardwicke, 1867 800, 1999 myeloma: Risk factors and effect on survival. Clin
2. Khorana AA: Malignancy, thrombosis and 10. Bryant AE: Biology and pathogenesis of Lymphoma 4:32-35, 2003
Trousseau: The case for an eponym. J Thromb thrombosis and procoagulant activity in invasive 17. Stein PD, Hull RD, Ghali WA, et al: Tracking
Haemost 1:2463-2465, 2003 infections caused by group A streptococci and Clos- the uptake of evidence: Two decades of hospital
3. Heit JA, Silverstein MD, Mohr DN, et al: Risk tridium perfringens. Clin Microbiol Rev 16:451-462, practice trends for diagnosing deep vein thrombosis
factors for deep vein thrombosis and pulmonary 2003 and pulmonary embolism. Arch Intern Med 163:
embolism: A population-based case-control study. 11. Keller TT, Mairuhu AT, de Kruif MD, et al:
1213-1219, 2003
Arch Intern Med 160:809-815, 2000 Infections and endothelial cells. Cardiovasc Res
18. Dhami MS, Bona RD, Calogero JA, et al:
4. Lee AY, Levine MN: Venous thromboembo- 60:40-48, 2003
Venous thromboembolism and high grade gliomas.
lism and cancer: Risks and outcomes. Circulation 12. Levi M, Keller TT, van Gorp E, et al: Infection
107:I17-I21, 2003 Thromb Haemost 70:393-396, 1993
and inflammation and the coagulation system. Car-
5. Ambrus JL, Ambrus CM, Mink IB, et al: Causes 19. Wun T, Law L, Harvey D, et al: Increased
diovasc Res 60:26-39, 2003
of death in cancer patients. J Med 6:61-64, 1975 incidence of symptomatic venous thrombosis in
13. Barbui T, Finazzi G, Grassi A, et al: Thrombosis
6. Levitan N, Dowlati A, Remick SC, et al: Rates in cancer patients treated with hematopoietic patients with cervical carcinoma treated with con-
of initial and recurrent thromboembolic disease growth factors: A meta-analysis—On behalf of the current chemotherapy, radiation, and erythropoietin.
among patients with malignancy versus those with- Subcommittee on Haemostasis and Malignancy of Cancer 98:1514-1520, 2003
out malignancy: Risk analysis using Medicare claims the Scientific and Standardization Committee of the 20. von Tempelhoff GF, Dietrich M, Niemann F, et
data. Medicine (Baltimore) 78:285-291, 1999 ISTH. Thromb Haemost 75:368-371, 1996 al: Blood coagulation and thrombosis in patients
7. Sallah S, Wan JY, Nguyen NP: Venous throm- 14. Stein PD, Hull RD, Kayali F, et al: Venous with ovarian malignancy. Thromb Haemost 77:456-
bosis in patients with solid tumors: Determination of thromboembolism according to age: The impact of 461, 1997
frequency and characteristics. Thromb Haemost 87: an aging population. Arch Intern Med 164:2260- 21. Elting LS, Escalante CP, Cooksley C, et al:
575-579, 2002 2265, 2004 Outcomes and cost of deep venous thrombosis
8. Rickles FR, Levine MN: Epidemiology of throm- 15. Solit DB, Morris M, Slovin S, et al: Clinical among patients with cancer. Arch Intern Med 164:
bosis in cancer. Acta Haematol 106:6-12, 2001 experience with intravenous estramustine phos- 1653-1661, 2004
www.jco.org 489
Downloaded from jco.ascopubs.org on December 27, 2015. For personal use only. No other uses without permission.
Copyright © 2006 American Society of Clinical Oncology. All rights reserved.
Khorana et al
22. Rickles FR, Patierno S, Fernandez PM: Tissue thromboembolism. N Engl J Med 343:1846-1850, prevention of thromboembolism in stage IV breast
factor, thrombin, and cancer. Chest 124:58S-68S, 2000 cancer. Lancet 343:886-889, 1994
2003 27. Kniffin WD Jr, Baron JA, Barrett J, et al: The 31. Kakkar AK, Levine M, Pinedo HM, et al: Ve-
23. Sahni A, Sporn LA, Francis CW: Potentiation epidemiology of diagnosed pulmonary embolism nous thrombosis in cancer patients: Insights from
of endothelial cell proliferation by fibrin(ogen)-bound and deep venous thrombosis in the elderly. Arch the FRONTLINE survey. Oncologist 8:381-388,
fibroblast growth factor-2. J Biol Chem 274:14936- Intern Med 154:861-866, 1994 2003
14941, 1999 28. Madsen K, Rosenman M, Hui S, et al: Value of 32. Zacharski LR, Henderson WG, Rickles FR, et
electronic data for model validation and refinement:
24. Palumbo JS, Kombrinck KW, Drew AF, et al: al: Effect of warfarin anticoagulation on survival in
Bacteremia risk in children with fever and neutrope-
Fibrinogen is an important determinant of the met- carcinoma of the lung, colon, head and neck, and
nia. J Pediatr Hematol Oncol 24:256-262, 2002
astatic potential of circulating tumor cells. Blood prostate: Final report of VA Cooperative Study #75.
29. Rosenman M, Madsen K, Hui S, et al: Model-
96:3302-3309, 2000 ing administrative outcomes in fever and neutrope- Cancer 53:2046-2052, 1984
25. Falanga A, Donati MB: Pathogenesis of nia: Clinical variables significantly influence length of 33. Khorana AA, Sahni A, Altland OD, et al:
thrombosis in patients with malignancy. Int J Hema- stay and hospital charges. J Pediatr Hematol Oncol Heparin inhibition of endothelial cell proliferation
tol 73:137-144, 2001 24:263-268, 2002 and organization is dependent on molecular
26. Sorensen HT, Mellemkjaer L, Olsen JH, et 30. Levine M, Hirsh J, Gent M, et al: Double-blind weight. Arterioscler Thromb Vasc Biol 23:2110-
al: Prognosis of cancers associated with venous randomised trial of a very-low-dose warfarin for 2115, 2003
■ ■ ■
Acknowledgment
We thank David Oakes, PhD, for critical comments and Susan Sullivan and Barbara Hartzog for editorial assistance.
Author Contributions