Current issues of ACP Journal Club are published in Annals of Internal Medicine


Concomitant VTE increased risks for mortality and hemorrhage in older patients with cancer, with risk varying by cancer type


ACP J Club. 2007 Sep-Oct;147:51. doi:10.7326/ACPJC-2007-147-2-051

Clinical Impact Ratings

Hematol/Thrombo: 5 stars

Oncology: 6 stars

Source Citation

Gross CP, Galusha DH, Krumholz HM. The impact of venous thromboembolism on risk of death or hemorrhage in older cancer patients. J Gen Intern Med. 2007;22:321-6. [PubMed ID: 17356962]



In older patients with a first diagnosis of cancer, how frequently is venous thromboembolism (VTE) diagnosed concomitantly? Does VTE increase risks for 1-year mortality and major hemorrhage?


Design: Inception cohort followed for 1 year.

Setting: United States.

Patients: 167 385 patients ≥ 67 years of age (median age 75 y, 52% men, 87% white) who were diagnosed with 1 of 10 cancer types (prostate, breast, bladder, uterus, lung, colorectal, lymphoma, pancreas, kidney, or ovary), had a malignant primary lesion with a known month of cancer diagnosis, and did not have a diagnosis of VTE 6 to 24 months before cancer diagnosis. Exclusion criteria included date of death before cancer diagnosis, ineligibility for Medicare Part A or B, and enrollment in a managed care plan within 2 years before cancer diagnosis.

Prognostic factors: Concomitant VTE (diagnosis between 6 mo before and 1 mo after cancer diagnosis), age, race, sex, cancer stage, histologic grade, chemotherapy, radiation, cancer-specific surgery, comorbid conditions, and socioeconomic status (SES).

Outcomes: Mortality and major hemorrhage (intracranial or gastrointestinal hemorrhage requiring hospital admission).

Main results

1.1% of patients had concomitant VTE. Concomitant VTE increased the risk for 1-year mortality for 8 of 10 cancer types (all but prostate and breast) (Table). Major hemorrhage occurred more frequently in patients with concomitant VTE than in those without VTE, both overall and for 8 of 10 cancer types (all but kidney and uterine) (Table).


In older patients with a first diagnosis of cancer, 1.1% had a concomitant diagnosis of venous thromboembolism (VTE). Concomitant VTE increased risks for 1-year mortality and major hemorrhage, with risk varying by cancer type.

Sources of funding: Claude D. Pepper Older Americans Independence Center at Yale and Boehringer-Ingelheim.

For correspondence: Dr. C.P. Gross, Yale University School of Medicine, New Haven, CT, USA. E-mail:

Table. Mortality and major hemorrhage at 1 year in older patients with a first diagnosis of cancer and concomitant VTE*

Cancer type Number of patients Hazard ratio (95% CI) for mortality associated with VTE† Major hemorrhage rates Excess hemorrhage rate in VTE patients (P value)
With concomitant VTE Without concomitant VTE
Breast 26 563 1.01 (0.77 to 1.33)‡ 10% 3.1% 6.9% (0.004)
Lung 32 348 1.16 (1.05 to 1.27) 13% 8.1% 4.5% (0.003)
Colorectal 29 101 1.19 (1.06 to 1.34) 26% 18% 7.9% (< 0.001)
Prostate 40 710 1.21 (0.96 to 1.52)‡ 8.5% 3.6% 4.9% (0.007)
Pancreas 6393 1.26 (1.06 to 1.49) 24% 16% 7.9% (0.02)
Ovary 3359 1.32 (1.03 to 1.68) 15% 8.0% 6.6% (0.033)
Bladder 11 063 1.43 (1.13 to 1.80) 13% 5.3% 7.9% (0.001)
Kidney 4141 1.43 (1.13 to 1.80) 11% 8.4% 2.1% (0.45)‡
Lymphoma 8022 1.63 (1.35 to 1.97) 22% 11% 12% (< 0.001)
Uterus 5685 1.96 (1.47 to 2.62) 1.9% 4.4% −2.5% (0.36)‡
Total 167 385 17% 7.9% 8.9% (< 0.001)

*VTE = venous thromboembolism; other abbreviations defined in Glossary. RRR, RRI, and CI calculated from data in article.
†Adjusted for age, sex, race, cancer characteristics, cancer treatments, comorbid conditions, and socioeconomic status.
‡Not significant.


Clinicians have likened decisions about anticoagulation to walking a tightrope—a misstep to one side increases the risk for bleeding and a misstep to the other increases the risk for thrombosis.

The study by Gross and colleagues emphasizes how often the fall into bleeding occurs. This well-done inception cohort study clarifies several key issues about older patients with VTE and cancer. It confirms that concomitant VTE is an ominous prognostic marker in older patients with some types of cancer and clarifies the mechanism for this observation.

Prognosis worsens when VTE accompanies cancer types usually detected at an advanced stage (e.g., pancreas) but not those detected at an early stage (e.g., breast or prostate). Similarly, patients with VTE and certain types of cancer had higher rates of major bleeding, although bleeding did not appear to be the cause of death. Therefore, it seems that the contribution of VTE to cancer mortality relates most to the stage of the underlying malignancy when VTE is diagnosed, rather than to VTE or its treatment. The study also highlights the high background risk for major bleeding (7.9%) in older patients with cancer, even in the absence of anticoagulants. This suggests that certain types of cancer should be risk factors in prediction tools for bleeding risk.

Some limitations bear note. As pharmacy data were not available, there was probably some contamination of the groups (i.e., it is not assured that all VTE patients received anticoagulants or that all non-VTE patients did not). However, the authors' correction for other common indications for anticoagulation mitigates this problem. Quality of warfarin management also could not be ascertained. Data were obtained before publication of guidelines (1) that suggest use of prolonged low-molecular-weight heparin products in lieu of warfarin. The impact of this recommendation on bleeding rates in older cancer patients therefore could not be addressed.

Much work remains to be done. Although we know that the risk for both thrombosis and hemorrhage is higher in older cancer patients than in patients without cancer, we still lack clear guidance on balancing these risks. At present, the results of Gross and colleagues' study should lead clinicians to carefully consider the risks and benefits of anticoagulants in patients with VTE receiving palliation for advanced cancer.

Scott M. Stevens, MD
C. Gregory Elliott, MD
LDS Hospital and the University of UtahSalt Lake City, USA


1. Buller HR, Agnelli G, Hull RD, et al. Antithrombotic therapy for venous thromboembolic disease: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 2004;126(3 Suppl):401-28S. [PubMed ID: 15383479]