Article Tools
Cancer-Related Complications
Article Tools
Venous Thromboembolism in Patients With Colorectal Cancer: Incidence and Effect on Survival
Abstract
To describe the incidence and outcomes associated with venous thromboembolism (VTE) among patients with colorectal cancer.
This was a retrospective analysis of all colorectal cancer patients diagnosed in California between 1993 and 1995 and 1997 to 1999. Principal outcomes were incident symptomatic VTE events and death. Associations between specific risk factors and principal outcomes were analyzed using Cox proportional hazards models.
Among 68,142 colorectal cancer patients, 50% were women, mean age was 70 ± 15 years, and approximately 70% underwent a major operation. The 2-year cumulative incidence of VTE was 2,100 patients (3.1%), with an incidence rate that decreased significantly over time from 5.0% (events/100 patient-years) in months 0 to 6 to 1.4% during months 7 to 12 to 0.6% during the second year. Significant predictors of VTE included metastatic stage (hazard ratio [HR] = 3.2; 95% CI, 2.8 to 3.8) and three or more comorbid conditions (HR = 2.0; 95% CI, 1.7 to 2.3). The risk of VTE was significantly reduced among Asians/Pacific Islanders (HR = 0.4; 95% CI, 0.3 to 0.5.) and patients who underwent an abdominal operation (HR = 0.4; 95% CI, 0.3 to 0.4). In risk-adjusted models, VTE was a significant predictor of death within 1 year of cancer diagnosis among patients with local- (HR = 1.8; 95% CI, 1.4 to 2.3) or regional-stage disease (HR = 1.5; 95% CI, 1.3 to 1.8) but not among patients with metastatic disease (HR = 1.1; 95% CI, 1.0 to 1.2).
The incidence of VTE among colorectal cancer patients was highest in the first 6 months after diagnosis and decreased rapidly thereafter. Metastatic disease and the number of medical comorbidities were the strongest predictors of VTE. Incident VTE reduced survival among patients with local or regional disease, suggesting that, in these patients, VTE may reflect the presence of a biologically more aggressive cancer.
Despite a long-recognized association between cancer and thrombosis,1-8 few data are available regarding the incidence of venous thromboembolism (VTE) among patients with specific types of cancer, such as colorectal cancer. A major impediment has been the absence of a method of identifying all incident cancer patients and determining the time of all subsequent VTE events among these patients. The best estimates of the incidence of VTE among colorectal cancer patients are based on the percentage of Medicare patients hospitalized for colorectal cancer who were rehospitalized for VTE or the percentage of colorectal patients who developed VTE after receiving chemotherapy.9-12
Recently, Chew et al13 used a population-based inception registry cohort linked with hospital discharge data to define the incidence and time course of VTE among patients with 12 common cancers. Patients with metastatic colorectal cancer had an incidence rate of approximately 4.3 VTE cases per 100 patient-years compared with 20.0 VTE cases per 100 patient-years in the highest risk group, who were patients with metastatic pancreatic cancer. However, this analysis did not look at issues specific to colorectal cancer, such as the effect of a major abdominal operation on the incidence of VTE or the relationship between cancer histology or location on the incidence of VTE. The objective of this study was to describe the incidence, time course, risk factors, and prognosis associated with VTE in a population-based cohort of patients with colorectal cancer.
This study was conducted using the following two merged databases: the California Cancer Registry, which includes 99% of all cancer patients diagnosed in California, and the California Patient Discharge Data Set, which contains linked medical-diagnostic and procedural data for all patients admitted to a public hospital in California. Coding uses International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) nomenclature. Approximately 5% of cancer registry patients could not be linked to the hospital discharge data set because they lacked a social security number.
The colon cancer cohort included all colorectal cancer patients 18 years or older who were diagnosed between January 1993 to December 1995 and January 1997 to December 1999. During this period, it was standard practice to admit patients with symptomatic VTE to a hospital for treatment. Federal hospitals (n = 14) were excluded because of the absence of any hospital discharge data. Registry information included basic demographics; the Surveillance, Epidemiology, and End Results cancer stage; date of diagnosis; and cancer histology. Tumor histologies were categorized into the following distinct groups: adenocarcinoma, mucinous adenocarcinoma, carcinoma in a polyp, undifferentiated carcinoma, and unspecified neoplasm. The date of cancer diagnosis was defined as the earlier of the following dates: the cancer registry diagnosis date or the date of a hospital admission that included an ICD-9-CM for primary or metastatic colon cancer.
The primary outcomes were the incidence of VTE and death. Death was determined using the linked California master death registry. Deep-vein thrombosis and pulmonary embolism were defined using previously validated ICD-9-CM codes (451.1x, 451.2, 451.81, 453.1, 453.2, 453.8, 453.9, and 415.1x) in the principal or a secondary position together with a hospital stay of 2 or more days, unless the patient died. Patients with superficial phlebitis or upper-extremity VTE were not included. Patients coded as having VTE before the cancer diagnosis were excluded.
The date of admission was considered the time of the VTE event for all patients admitted with a principal diagnosis of VTE. Among patients with a secondary diagnosis of VTE and an associated procedure code (eg, venous ultrasound or inferior vena cava filter), the VTE date was the procedure date. For patients with a secondary diagnosis of VTE and no test date, the VTE diagnosis date assigned was the median day of hospitalization.14
All first-time VTE patients were classified in a hierarchical fashion into the following eight groups based on the presence or absence of specific risk factors defined using specific ICD-9-CM codes: (1) major abdominal surgery–associated VTE (< 61 days, see Appendix); (2) other malignancy–associated VTE; (3) trauma-associated VTE (< 61 days); (4) colorectal biopsy surgery–associated VTE (see Appendix); (5) other surgery–associated VTE (< 61 days); (6) recent medical hospitalization–associated VTE (within 61 days of hospital stay of ≥ 4 days); (7) concurrent medical hospitalization–associated VTE; and (8) spontaneous VTE (all remaining patients admitted for VTE).
The effect of chronic comorbid medical conditions was analyzed using a variation of the Elixhauser Comorbidity Index.15 Terms indicating the presence of cancer (tumor, metastatic, and lymphoma) or more acute conditions (electrolyte disturbance, arrhythmia, and coagulopathy) were excluded. The index hospitalization and all hospitalizations less than 2 years before the cancer diagnosis were used to generate the comorbidity index.16
Incidence rates were calculated as both cumulative incidence and as person-time (events/100 patient-years). Differences in incidence were tested using the normal approximation to the binomial distribution. Cox proportional hazards models were used to analyze the effect of specified risk factors on the outcomes of VTE or death within 1 year of cancer diagnosis. Proportionality assumptions of the models were checked and were met by the data. Kaplan-Meier plots were generated to illustrate the effect of VTE on survival, matching four patients without VTE to each patient with VTE based on survival from the cancer diagnosis date to the VTE diagnosis date, age (within 2 years), race, and cancer stage. Cancer patients who never underwent a major operation were matched (1:1) with patients who underwent an operation based on survival from the cancer diagnosis to the day of the operation, age, race, and initial cancer stage. Differences in survival among comparison cohorts were assessed using the log-rank test. Analyses were performed using SAS (SAS Institute, Cary, NC), S-plus (Version 3.3; Statistical Sciences, Seattle, WA), or SISA (http://home.clara.net/sisa/smr.htm; Quantitative Skill, Hilversum, the Netherlands). Given the large sample size, in multivariate models and comparisons of survival, P < .0001 was used to define significance. This study was approved by the California Health and Welfare Agency Committee for the Protection of Human Subjects and by the University of California Davis (Davis, CA) Human Subjects Committee.
During the 6-year time period, there were 68,142 patients diagnosed with colorectal cancer; 50% were male, and 50% were female, with an average age of 70 ± 15 years. Whites made up 75% of the cohort. Most patients (73%) had local or regional disease; adenocarcinoma was the most prevalent histologic type (77%), rectosigmoid location was the most prevalent tumor site (51%), and almost 70% of patients underwent a major colorectal operation within days to weeks of the date of diagnosis.
The 2-year cumulative incidence of VTE after cancer diagnosis was 2,100 patients (3.1%). The incidence rate was 5.0% (events/100 patient-years) during months 0 to 6 after cancer diagnosis, 1.4% during the next 6-month period, and 0.6% during the second year of follow-up. Table 1 illustrates the relationship between various clinical and demographic features and the incidence of VTE. In bivariate analyses, compared with whites, the cumulative incidence of VTE was significantly lower among Hispanics (P < .02) and Asians/Pacific Islanders (P < .001). Patients with mucinous adenocarcinoma had a slightly higher incidence of VTE compared with patients with adenocarcinoma (P < .01.) The incidence of VTE increased significantly with advancing stage, with a 2-year cumulative incidence of 4.7% among patients with metastatic disease compared with 1.8% among patients with local-stage disease (P < .0001).
Rectosigmoid cancer was associated with a slightly but significantly lower incidence of VTE compared with cancers in the ascending, transverse, or descending colon. A total of 1,016 patients (1.5%) underwent partial hepatectomy, presumably for isolated hepatic metastasis. Their 2-year cumulative incidence of VTE was 3.6% v 3.1% for the entire cohort (P = .4).
Table 2 lists a hierarchical classification of all VTE events diagnosed within 2 years of colorectal cancer diagnosis. Approximately 30% of the VTE patients were diagnosed within 2 months of a major abdominal operation, 8% were diagnosed after another operation, 9% were diagnosed during or within 2 months of a medical hospitalization, and 42% had no associated provoking risk factor other than the cancer diagnosis.
Figure 1 is a Kaplan-Meier plot of the incidence of VTE stratified by cancer stage, which shows a strong relationship between cancer stage and the incidence of VTE. In this plot, the day of diagnosis of VTE was assigned as day 0 for all patients diagnosed during the index hospitalization. For all stages of cancer, the incidence rate decreased over time. Figure 2 shows the incidence of VTE among the patients who underwent a major operation within 2 months of cancer diagnosis compared with a matched sample of patients who did not undergo a major operation.
The results of multivariate models are listed in Table 3. In the model predicting development of VTE within 1 year, increased risk was associated with regional stage (hazard ratio [HR] = 2.1; 95% CI, 1.8 to 2.4) or metastatic stage (HR = 3.2; 95% CI, 2.8 to 3.8) and the number of chronic comorbid conditions. Asians/Pacific Islanders (HR = 0.4; 95% CI, 0.3 to 0.5) and patients undergoing a major abdominal operation (HR = 0.4; 95% CI, 0.3 to 0.4) had a decreased risk.
Significant predictors of death within 1 year of the cancer diagnosis included advancing age, regional or metastatic disease, histology showing undifferentiated carcinoma, increasing number of comorbid conditions, and a diagnosis of VTE. Figure 3A shows a Kaplan-Meier plot comparing the incidence of death after VTE to a matched cohort without VTE. Figures 3B, 3C, and 3D are similar Kaplan-Meier plots stratified by initial cancer stage. Differences in the incidence of death over time were largest among the patients with local- or regional-stage cancer, with no significant difference among the patients who had metastatic disease. Differences in mortality were apparent within 50 days of the day of the VTE diagnosis and showed little change after 180 days. In risk-adjusted proportional hazards models stratified by initial cancer stage, the risk of death among patients diagnosed with VTE (v no VTE) was 1.8-fold higher among patients with local-stage disease (HR = 1.8; 95% CI, 1.4 to 2.3; P < .0001), 1.5-fold higher among patients with regional-stage disease (HR = 1.5; 95% CI, 1.3 to 1.7; P < .0001), and not significantly different among patients with metastatic disease (HR = 1.1; 95% CI, 1.0 to 1.2; P = .26).
This population-based study provided a robust analysis of the epidemiology of VTE among patients with newly diagnosed colorectal cancer. Essentially all colorectal cancer patients diagnosed in California during a 6-year period were identified using the state cancer registry, and nearly all subsequent symptomatic VTE events that were diagnosed in California hospitals were then detected. Major findings included a significant decrease in the incidence rate of VTE over time, a strong relationship between the presence of metastatic disease at the time of diagnosis and the incidence of VTE, and a significant reduction in survival among the patients diagnosed with VTE, which was restricted to the patients who initially had local- or regional-stage cancer. Other significant findings included a decreased risk of VTE among Asians/Pacific Islanders, a modestly lower incidence of VTE among patients with rectosigmoid cancer, and both a lower incidence of VTE and improved survival among the patients who underwent a major abdominal operation. In the multivariate analysis, there was no significant difference in the incidence of VTE based on tumor histology.
The incidence rate of VTE was highest in the first 6 months after diagnosis of cancer, which is a finding recently reported in a case-control study of patients with acute VTE.17 Calculated as events per 100 patient-years, the incidence rate decreased from 5% during the first 6 months after diagnosis to 1.4% during the next 6-month period to 0.6% during the second year of follow-up. This decrease in the incidence rate of VTE occurred despite progression of the cancer in a substantial percentage of patients, as evidenced by a continued high mortality. For example, among the 26,740 patients with regional-stage disease, there were 841 VTE events (3.1%) diagnosed within 1 year of the cancer diagnosis, during which time 3,933 patients (14.7%) died. However, during the second year of follow-up, there were only 122 VTE events (0.5%), despite 3,098 additional deaths (representing 13.6% of the patients alive after 1 year.) This finding suggests that thrombosis is related more to the initial biologic properties of the cancer than to spread of the disease.
The incidence of VTE was higher among patients who never underwent a major abdominal operation, probably because these patients had advanced inoperable cancer or significant underlying comorbidities that precluded an operation. In the multivariate model to predict development of VTE within 1 year, undergoing a major abdominal operation was associated with a 60% relative risk reduction even after adjusting for the presence of medical comorbidity. The lower long-term incidence of VTE noted 2 to 12 months after operation might reflect the curative resection of cancer in a substantial percentage of the patients. Although these results are provocative, we stress that any comparison of the outcomes between patients undergoing an operation and those without an operation should be interpreted with great caution because of obvious selection bias associated with the determination of which patients would be best served by an operation.
The lower incidence of VTE among the patients who had rectosigmoid disease was unexpected because of the more extensive surgery associated with abdominal-perineal resection and resection of lymph nodes near the pelvic veins and the fact that local recurrence frequently involves the pelvic sidewall. During the time period analyzed, standard practice for patients with regional disease of the rectum was postoperative chemotherapy and radiation, which should have increased the incidence of VTE even further. Possibly, more aggressive thromboprophylaxis was administered to these patients, but this is purely speculative.
Diagnosis of a VTE had a significant adverse effect on survival, which is a finding that has previously been reported.3,4,8,13,18 The 2-year cumulative incidence of death among patients who never developed VTE was 35% v 52% among the patients who developed VTE (P < .0001). In a multivariate model to predict death within 1 year, a diagnosis of VTE was associated with a 20% higher risk of dying. Interestingly, when analyzed by cancer stage, patients with local-stage cancer and VTE were 1.8 times more likely to die than patients without VTE, and patients with regional disease who developed VTE were 1.5 times more likely to die. However, among patients with an initial diagnosis of metastatic disease, there was no difference in survival among the patients who either had or did not have a VTE. The effect of VTE on survival among the patients with local- or regional-stage disease was most pronounced in the first 50 to 100 days after VTE diagnosis. These findings have not been previously reported. In the study by Chew et al,13 development of VTE was associated with reduced survival even among patients with metastatic disease, but this analysis did not adjust for comorbidities, tumor histology, location, or operations.
There are several possible explanations for the finding of reduced survival restricted to the VTE patients who had local- or regional-stage disease initially. If development of VTE in a portion of these patients reflects the presence of a biologically more aggressive cancer that, in turn, leads to reduced survival, this effect is much more likely to be measurable among patients with a better prognosis. The reduction in survival among the patients with local- or regional-stage cancer who developed VTE was approximately 10%. The fact that this was not seen in the patients who had metastatic-stage cancer initially probably reflects their poor prognosis. Other possible explanations for the reduced survival among the VTE patients include confounding as a result of the presence of significant unmeasured comorbidity, incorrect initial staging, and death as a result of bleeding or thromboembolic complications.
Nevertheless, the finding of reduced survival among local- or regional-stage patients who developed VTE is highly provocative in light of the recent posthoc analysis by Lee et al,19 who reported that 6 months of treatment using a low molecular weight heparin had a survival advantage compared with warfarin among cancer patients with VTE who had local- or regional-stage disease but not metastatic disease. Certainly, the findings of the present study suggest that some of the patients with limited-stage colon cancer who develop VTE are in some way unique or different. These data are consistent with recent reports suggesting that thrombin generation is associated with increased mortality in patients with cancer.20-22 Thus, in at least a subgroup of patients diagnosed with local- or limited-stage colon cancer, development of VTE within a year may reflect cancer-associated enhanced thrombin generation, with reduced survival in these patients because the cancer is biologically more aggressive.
The major limitation of this retrospective study was the use of administrative data to determine the incidence of VTE, rather than prospectively collected objectively defined data. However, prior validation studies have shown that the codes that were used to define VTE have a positive predictive value of approximately 90%.14,23,24 Moreover, the findings from studies that rely on administrative data can frequently be corroborated based on the results of prospective clinical trials or cohort studies. For example, in the current study, there was a steady increase in the relative risk of death as the number of comorbid conditions increased, which is a finding similar to the results of a recent prospective cohort study of cancer patients.25 Another limitation was the absence of detailed information about the use of chemotherapy, which was not reliably captured by the registry. Finally, the registry lacked detailed information regarding the use of antithrombotic prophylaxis. However, absence of information about thromboprophylaxis would probably only affect the analysis of postoperative VTE events. During the time analyzed in this study, primary VTE thromboprophylaxis was rarely administered, and extended prophylaxis after a medical hospitalization or surgical operation was not routine.26
Among patients with colorectal cancer, approximately 3% developed VTE within 2 years, and the incidence of VTE was highest in the first few months after cancer diagnosis. Significant risk factors associated with the development of VTE included advancing cancer stage at the time of diagnosis and the presence of comorbid medical conditions. Asians/Pacific Islanders and patients who underwent a major abdominal operation exhibited a decreased risk of developing VTE. Development of a VTE was a significant risk factor for death but only among the patients with local- and regional-stage disease. Further studies are needed to determine why survival is reduced among these patients and whether extended use of low molecular weight heparin leads to improved survival in these patients.
ICD-9 Codes Used to Categorize Procedures in the Colorectal Cancer Cohort ICD-9 Codes of Operations included as Major Abdominal SurgeryTransabdominal endoscopy of large intestine (laparoscopy) ICD-9 Codes of Operations included As Major Abdominal Surgery 458Total intra-abdominal colectomy 485Abdominoperineal resection of rectum 4500Small bowel incision NOS 4501Incision of duodenum 4502Other incision of small intestine 4503Incision of large intestine (enterotomy) 4550Isolation of intestinal segment NOS 4551Isolation of segment of small intestine 4552Isolation of segment of large intestine 4560Other excision of small intestine 4561Multiple segmental resection of small intestine 4562Other partial resection of small intestine (ileectomy) 4563Total removal small intestine 4570Partial excision large intestine 4571Multiple segmental resection of large intestine 4572Cecectomy 4573Right hemicolectomy 4574Resection transverse colon 4575Left hemicolectomy 4576Sigmoidectomy 4579Other partial excision of large intestine 4590Intestinal anastomosis NOS 4591Small to small intestinal anastomosis 4592Anastomosis of small intestine to rectal stump 4593Other small to large anastomosis 4594Large to large intestine anastomosis 4595Anastomosis to anus 4600Exteriorization of intestine and loop enterostomy 4601Loop ileostomy 4602Resection of exteriorized segment of small intestine 4603Loop colostomy (continued on following page)
The author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
| Authors | Employment | Leadership | Consultant | Stock | Honoraria | Research Funds | Testimony | Other |
|---|---|---|---|---|---|---|---|---|
| Richard H. White | Boehringer Ingelheim (A); Agenix (A) | Pharmion (A); Sanofi Aventis (A); Agen (A) |
Dollar Amount Codes (A) < $10,000 (B) $10,000-99,999 (C) ≥ $100,000 (N/R) Not Required
Conception and design: Ted Wun, Helen K. Chew, Richard H. White
Financial support: Helen K. Chew, Richard H. White
Administrative support: Allison Alcalay
Provision of study materials or patients: Helen K. Chew, Ted Wun, Richard H. White
Collection and assembly of data: Allison Alcalay, Helen K. Chew, Hong Zhou, Danielle Harvey, Richard H. White
Data analysis and interpretation: Allison Alcalay, Ted Wun, Vijay Khatri, Helen K. Chew, Hong Zhou, Danielle Harvey, Richard H. White
Manuscript writing: Allison Alcalay, Vijay Khatri, Richard H. White, Ted Wun
Final approval of manuscript: Allison Alcalay, Ted Wun, Vijay Khatri, Helen K. Chew, Danielle Harvey, Richard H. White
Fig 1. Kaplan-Meier plot of the incidence of venous thromboembolism (VTE) after colorectal cancer diagnosis stratified by stage.
Fig 2. Comparison of the incidence of venous thromboembolism (VTE) among patients who underwent a major operation within 2 months of cancer diagnosis and matched patients who never underwent an operation.
Fig 3. Kaplan-Meier plots comparing the incidence of death after venous thromboembolism (VTE) with the incidence of death in matched patients who never developed VTE. (A) Entire cohort (log-rank χ2 = 59.9; P < .0001) and patients with (B) local-stage (log-rank χ2 = 37.1; P < .0001), (C) regional-stage (log-rank χ2 = 47.6; P < .0001), and (D) metastatic-stage disease (log-rank χ2 = 7.4; P < .007).
|
| Characteristic | Patients | Rate of VTE at 6 Months (per 100 patient-years)* | 2-Year Cumulative Incidence of VTE | 2-Year Incidence of Death (%) | ||||
|---|---|---|---|---|---|---|---|---|
| No. | % | No. | Row % | |||||
| Total patients | 68,142 | 5.0 | 2,100 | 3.1 | 35.5 | |||
| Age groups | ||||||||
| ≤ 50 years, referent | 6,109 | 9 | 3.6 | 157 | 2.6 | 27.4 | ||
| 51-60 years | 9,084 | 13 | 3.8 | 253 | 2.8† | 26.2 | ||
| 61-70 years | 16,455 | 24 | 4.9 | 555 | 3.4‡ | 28.3 | ||
| 71-80 years | 21,382 | 31 | 5.7 | 738 | 3.5§ | 35.3 | ||
| ≥ 81 years | 15,112 | 22 | 5.4 | 397 | 2.6† | 52.7 | ||
| Sex‖ | ||||||||
| Male, referent | 33,932 | 50 | 4.8 | 1,026 | 3.0 | 34.7 | ||
| Female | 34,202 | 50 | 5.1 | 1,074 | 3.1† | 36.4 | ||
| Race/ethnicity | ||||||||
| White, referent | 50,941 | 75 | 5.3 | 1,668 | 3.3 | 36.0 | ||
| African American | 4,385 | 6 | 6.2 | 162 | 3.7† | 41.2 | ||
| Hispanic | 6,874 | 10 | 3.9 | 188 | 2.7‡ | 33.7 | ||
| Asian/Pacific Islander | 5,557 | 8 | 2.0 | 73 | 1.3§ | 30.2 | ||
| Other | 385 | 0.6 | 4.0 | 9 | 2.3† | 22.3 | ||
| Histologic group | ||||||||
| Adenocarcinoma, referent | 45,476 | 67 | 5.2 | 1,514 | 3.3 | 36.3 | ||
| Mucin adenocarcinoma | 6,660 | 10 | 6.5 | 263 | 3.9‡ | 28.6 | ||
| Carcinoma in a polyp | 12,638 | 19 | 3.1 | 251 | 2.0§ | 18.0 | ||
| Undifferentiated carcinoma | 3,368 | 5 | 7.8 | 72 | 2.1§ | 84.5 | ||
| Stage | ||||||||
| Local, referent | 23,215 | 34 | 2.7 | 411 | 1.8 | 14.5 | ||
| Regional | 26,740 | 39 | 5.3 | 963 | 3.6§ | 26.3 | ||
| Metastatic | 13,228 | 19 | 9.6 | 627 | 4.7§ | 79.2 | ||
| Unclassified | 4,959 | 7 | 4.2 | 99 | 2.0† | 67.3 | ||
| Surgery | ||||||||
| Yes, referent | 46,598 | 68 | 4.0 | 1,281 | 2.7 | 28.1 | ||
| No | 21,544 | 32 | 7.3 | 819 | 3.8§ | 51.7 | ||
| Location | ||||||||
| Colon, referent | 19,160 | 28 | 5.8 | 649 | 3.4 | 36.2 | ||
| Cecum | 11,373 | 17 | 6.3 | 444 | 3.9‡ | 39.2 | ||
| Rectosigmoid | 34,515 | 51 | 4.1 | 945 | 2.7§ | 30.4 | ||
| Unclassified | 3,094 | 5 | 5.8 | 62 | 2.0§ | 75.2 | ||
Abbreviation: VTE, venous thromboembolism.
*Rate of VTE per 100 patient-years in the first 6 months after diagnosis of cancer.
†P = not significant.
‡P < .05.
§P < .0001.
‖For eight patients, sex was unknown or classified as hermaphrodite.
|
| VTE Category | No. of Patients (2 years) | % of Total Patients With VTE |
|---|---|---|
| Major abdominal surgery, < 61 days | 642 | 30.5 |
| Other, noncolorectal malignancy | 45 | 2.1 |
| Trauma, < 61 days | 15 | 0.7 |
| Colorectal biopsy, < 61 days | 12 | 0.6 |
| Other surgery, < 61 days | 158 | 7.5 |
| Recent medical hospitalization, < 2 months | 185 | 8.8 |
| Concurrent medical hospitalization | 162 | 7.7 |
| Spontaneous | 881 | 42.0 |
Abbreviation: VTE, venous thromboembolism.
|
| Variable | Hazard Ratio of VTE < 1 Year | 95% CI | Hazard Ratio of Death < 1 Year | 95% CI |
|---|---|---|---|---|
| Men v women | 1.0* | 0.9 to 1.1 | 0.9† | 0.9 to 1.0 |
| Race/ethnicity v white | ||||
| African American | 0.9* | 0.8 to 1.1 | 1.0* | 1.0 to 1.1 |
| Hispanic | 0.8† | 0.7 to 1.0 | 1.0* | 0.9 to 1.0 |
| Asian/Pacific Islander | 0.4† | 0.3 to 0.5 | 0.8‡ | 0.8 to 0.9 |
| Age v ≤ 50 years | ||||
| 51-60 years | 1.1* | 0.9 to 1.3 | 1.1* | 1.0 to 1.2 |
| 61-70 years | 1.3‡ | 1.0 to 1.5 | 1.3† | 1.2 to 1.5 |
| 71-80 years | 1.2‡ | 1.0 to 1.5 | 1.9† | 1.7 to 2.0 |
| ≥ 81 years | 0.9* | 0.7 to 1.1 | 2.9† | 2.7 to 3.1 |
| Stage v local | ||||
| Regional | 2.1† | 1.8 to 2.4 | 1.6† | 1.5 to 1.7 |
| Metastatic | 3.2† | 2.8 to 3.8 | 8.3† | 7.8 to 8.7 |
| Unknown | 1.3* | 1.0 to 1.6 | 4.2† | 3.9 to 4.5 |
| Histology v adenocarcinoma | ||||
| Mucinous | 1.1* | 1.0 to 1.3 | 1.0* | 0.9 to 1.0 |
| Undifferentiated or NOS | 1.1* | 0.8 to 1.4 | 2.0† | 1.9 to 2.1 |
| Polyp | 0.8‡ | 0.7 to 1.0 | 0.8† | 0.7 to 0.8 |
| Location v colon | ||||
| Cecum | 1.1* | 1.0 to 1.3 | 1.0* | 0.9 to 1.0 |
| Rectosigmoid | 0.9‡ | 0.8 to 1.0 | 0.8† | 0.8 to 0.9 |
| Other | 0.7‡ | 0.5 to 0.9 | 1.3† | 1.3 to 1.4 |
| No. of comorbidities v 0 | ||||
| 1 | 1.2‡ | 1.1 to 1.4 | 1.2† | 1.2 to 1.3 |
| 2 | 1.5† | 1.3 to 1.8 | 1.5† | 1.4 to 1.6 |
| ≥ 3 | 2.0† | 1.7 to 2.3 | 2.1† | 2.0 to 2.2 |
| Major surgery v none | 0.4† | 0.3 to 0.4 | 0.48† | 0.46 to 0.5 |
| VTE in first year | — | — | 1.2† | 1.1 to 1.3 |
Abbreviations: VTE, venous thromboembolism; NOS, not otherwise specified.
*P = not significant.
†P < .0001.
‡P < .05.
Supported in part by National Institutes of Health Grant No. 1-RO3- CA99527-01.
Presented in part at the 20th Congress of the International Society of Thrombosis and Haemostasis, Sydney, Australia, August 6-12, 2005.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
| 1. | Thompson CM, Rodgers LR: Analysis of the autopsy records of 157 cases of carcinoma of the pancreas with particular reference to the incidence of thromboembolism. Am J Med Sci 223::469,1952-478, Crossref, Medline, Google Scholar |
| 2. | Shen VS, Pollak EW: Fatal pulmonary embolism in cancer patients: Is heparin prophylaxis justified? South Med J 73::841,1980-843, Crossref, Medline, Google Scholar |
| 3. | Prandoni P, Lensing AW, Cogo A, et al: The long-term clinical course of acute deep venous thrombosis. Ann Intern Med 125::1,1996-7, Crossref, Medline, Google Scholar |
| 4. | Sorensen HT, Mellemkjaer L, Olsen JH, et al: Prognosis of cancers associated with venous thromboembolism. N Engl J Med 343::1846,2000-1850, Crossref, Medline, Google Scholar |
| 5. | Heit JA, Silverstein MD, Mohr DN, et al: Risk factors for deep vein thrombosis and pulmonary embolism: A population-based case-control study. Arch Intern Med 160::809,2000-815, Crossref, Medline, Google Scholar |
| 6. | Rickles FR, Levine MN: Epidemiology of thrombosis in cancer. Acta Haematol 106::6,2001-12, Crossref, Medline, Google Scholar |
| 7. | Joung S, Robinson B: Venous thromboembolism in cancer patients in Christchurch, 1995-1999. N Z Med J 115::257,2002-260, Medline, Google Scholar |
| 8. | Lee AY, Levine MN: Venous thromboembolism and cancer: Risks and outcomes. Circulation 107:I17-I21, 2003 Google Scholar |
| 9. | Levitan N, Dowlati A, Remick SC, et al: Rates of initial and recurrent thromboembolic disease among patients with malignancy versus those without malignancy: Risk analysis using Medicare claims data. Medicine (Baltimore) 78::285,1999-291, Crossref, Medline, Google Scholar |
| 10. | Thodiyil PA, Kakkar AK: Variation in relative risk of venous thromboembolism in different cancers. Thromb Haemost 87::1076,2002-1077, Crossref, Medline, Google Scholar |
| 11. | Otten HM, Mathijssen J, ten Cate H, et al: Symptomatic venous thromboembolism in cancer patients treated with chemotherapy: An underestimated phenomenon. Arch Intern Med 164::190,2004-194, Crossref, Medline, Google Scholar |
| 12. | Grem JL, McAtee N, Murphy RF, et al: Phase I and pharmacokinetic study of recombinant human granulocyte-macrophage colony-stimulating factor given in combination with fluorouracil plus calcium leucovorin in metastatic gastrointestinal adenocarcinoma. J Clin Oncol 12::560,1994-568, Link, Google Scholar |
| 13. | Chew H, Wun T, Harvey D, et al: The incidence of venous thromboembolism and its effect on survival among patients with common cancers. Arch Intern Med (in press) Google Scholar |
| 14. | White RH, Romano PS, Zhou H, et al: Incidence and time course of thromboembolic outcomes following total hip or knee arthroplasty. Arch Intern Med 158::1525,1998-1531, Crossref, Medline, Google Scholar |
| 15. | Elixhauser A, Steiner C, Harris DR, et al: Comorbidity measures for use with administrative data. Med Care 36::8,1998-27, Crossref, Medline, Google Scholar |
| 16. | Stukenborg GJ, Wagner DP, Connors AF Jr: Comparison of the performance of two comorbidity measures, with and without information from prior hospitalizations. Med Care 39::727,2001-739, Crossref, Medline, Google Scholar |
| 17. | Blom JW, Doggen CJ, Osanto S, et al: Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 293::715,2005-722, Crossref, Medline, Google Scholar |
| 18. | Carson JL, Kelley MA, Duff A, et al: The clinical course of pulmonary embolism. N Engl J Med 326::1240,1992-1245, Crossref, Medline, Google Scholar |
| 19. | Lee AY, Rickles FR, Julian JA, et al: Randomized comparison of low molecular weight heparin and coumarin derivatives on the survival of patients with cancer and venous thromboembolism. J Clin Oncol 23::2123,2005-2129, Link, Google Scholar |
| 20. | den Ouden M, Ubachs JM, Stoot JE, et al: Thrombin-antithrombin III and D-dimer plasma levels in patients with benign or malignant ovarian tumours. Scand J Clin Lab Invest 58::555,1998-559, Crossref, Medline, Google Scholar |
| 21. | Beer JH, Haeberli A, Vogt A, et al: Coagulation markers predict survival in cancer patients. Thromb Haemost 88::745,2002-749, Crossref, Medline, Google Scholar |
| 22. | Rickles FR, Patierno S, Fernandez PM: Tissue factor, thrombin, and cancer. Chest 124::58S,2003-68S, (suppl 3) Crossref, Medline, Google Scholar |
| 23. | White RH, Brickner L, Scannell K: ICD-9-CM codes poorly identified venous thromboembolism during pregnancy. J Clin Epidemiol 57::985,2004-988, Crossref, Medline, Google Scholar |
| 24. | White RH, Gettner S, Newman JM, et al: Predictors of rehospitalization for symptomatic venous thromboembolism after total hip arthroplasty. N Engl J Med 343::1758,2000-1764, Crossref, Medline, Google Scholar |
| 25. | Piccirillo JF, Tierney RM, Costas I, et al: Prognostic importance of comorbidity in a hospital-based cancer registry. JAMA 291::2441,2004-2447, Crossref, Medline, Google Scholar |
| 26. | Levine M, Gent M, Hirsh J, et al: A comparison of low-molecular-weight heparin administered primarily at home with unfractionated heparin administered in the hospital for proximal deep-vein thrombosis. N Engl J Med 334::677,1996-681, Crossref, Medline, Google Scholar |
