Statins Decrease the Occurrence of Venous Thromboembolism in Patients with Cancer

Materials and Methods 

We conducted a retrospective case-control study. All patients with a history of solid organ tumor, such as breast, colon, lung, prostate, stomach, esophagus, pancreas, ovary, kidney, and brain cancer (Table 1), who were admitted to Albert Einstein Medical Center between October 2004 and September 2007 were eligible. The occurrence of venous thromboembolism and concurrent use of a statin within 2 months of admission were recorded. Statin use included simvastatin (Zocor; Merck & Co, Inc), atorvastatin (Lipitor; Pfizer U.S. Pharmaceuticals, New York, NY), rosuvastatin (Crestor; AstraZeneca Pharmaceuticals, LP), lovastatin (Mevacor; Merck & Co, Inc), pravastatin (Pravachol; Bristol-Myers Squibb, New York, NY), and fluvastatin (Lescol, Abbott Laboratories, Abbott Park, Ill). Dosage of statin use also was recorded. Possible risk factors for venous thromboembolism and comorbidities were assessed. Patients who never used statins or had used them for less than 2 months were assigned to the control group. All patients included had at least 2 hospital admissions.

Table 1. Type of Solid Organ Tumor

	Type of Solid Organ Tumor	No. of Patients (N=740)
	Breast cancer	170(23%)
	Lung cancer	95(13%)
	Colon cancer	55(8%)
	Prostate cancer	160(21%)
	Stomach cancer	20(3%)
	Esophageal cancer	45(6%)
	Pancreatic cancer	56(8%)
	Ovarian cancer	50(7%)
	Kidney cancer	60(8%)
	Brain cancer	29(4%)

In total, 1194 patients with a diagnosed solid organ tumor were identified (Figure 1). However, 454 patients were excluded for the following reasons: advanced age, benign tumor, other types of malignancy, incomplete medical record, or currently receiving treatment with anticoagulation therapy (oral warfarin or intravenous heparin). A total of 740 patients were included in this study. The diagnosis of deep vein thrombosis was confirmed by Doppler ultrasound with the single criteria of vein compressibility with the transducer probe. The femoral and popliteal veins were evaluated for noncompressibility. Pulmonary embolism was diagnosed by computed tomography pulmonary angiography with the presence of segmental or subsegmental emboli in pulmonary veins or high probability on ventilation-perfusion scan with high clinical probability.

View Large Image Download to PowerPoint

Figure 1. Study population.

All data were collected from the electronic medical record system. Demographic data and laboratory values were collected, including age, gender, and race. In addition, data were collected on comorbidities, risk factors for deep vein thrombosis or pulmonary embolism, additional treatments (diagnosis of diabetes, hypertension, immobilization, documented metastatic disease, current use of chemotherapy, history of smoking, use of aspirin, statin use at the time of admission), and whether or not the patient had a deep vein thrombosis or pulmonary embolism. Patients who had either a deep vein thrombosis or a pulmonary embolism were included in the venous thromboembolism group.

Results 

The mean age of the study sample was 65 years (standard deviation=13). Fifty-two percent of the patients were women, and 76% were African American. The overall frequency of venous thromboembolism was 18% (N=132), 31% of patients had a history of dyslipidemia, and 26% (N=194) were receiving statins. Baseline characteristics of the study sample comparing the statin user group with the control group are shown in Table 2. The distribution of the use of statins is listed in Table 3.

Table 2. Baseline Characteristics

		Statin User (%) N=194	Control Group (%) N=546	Total N=740
	Age	66 y	65 y	65 y
	Gender
	Female	97(50%)	289(53%)	386(52%)
	Male	97(50%)	257(47%)	354(48%)
	Race
	African American	144(74%)	418(76%)	562(76%)
	Caucasian	37(19%)	92(17%)	129(17%)
	Others	13(7%)	36(7%)	49(7%)
	Diabetes	82(42%)	110(20%)	192(26%)
	Hypertension	170(87%)	322(59%)	492(66%)
	Immobilization	33(25%)	71(12%)	104(14%)
	Metastatic disease	53(40%)	131(21%)	184(24%)
	Current chemotherapy	37(28%)	96(16%)	133(18%)
	Smoking	28(21%)	183(30%)	211(29%)
	Use of aspirin	19(14%)	138(23%)	157(21%)

Table 3. Summary of Statins Used

	Type of Statins Used	Standard Dose (<40 mg) (N=184)	High Dose (>40 mg) (N=10)	Total No. of Patients (N=194)
	Simvastatin	50	2	52(27%)
	Atorvastatin	103	3	106(55%)
	Lovastatin	14	2	16(8%)
	Pravastatin	10	2	12(6%)
	Rosuvastatin/fluvastatin	7	1	8(4%)

The average duration of the follow-up period for statin users was 10.2 months (range, 2-41 months). Among statin users, the majority of patients (95%) were using a standard dose of statins (up to 40 mg per day of any statins). In a comparison of 2 frequently prescribed statins, there were fewer occurrences of venous thromboembolism with simvastatin compared with atorvastatin (7.7% in the simvastatin group and 8.5% in the atorvastatin group, P=.029).

Having a history of diabetes and hypertension did not affect the odds of having a venous thromboembolism (OR 0.94, 95% confidence interval [CI], 0.61-1.45; OR 0.82, 95% CI, 0.56-1.22, respectively). Immobilization, documented metastatic disease, and current use of chemotherapy statistically significantly increased the odds of developing venous thromboembolism (OR 2.52, 95% CI, 1.58-4.01; OR 2.44, 95% CI, 1.64-3.63; OR 2.07, 95% CI, 1.34-3.22, respectively). Smoking, use of aspirin, and use of statins were related to a decreased odds of developing a venous thromboembolism (OR 0.62, 95% CI, 0.39-0.98; OR 0.57, 95% CI, 0.34-0.96; OR 0.33, 95% CI, 0.19-0.57, respectively) (Table 4 and Figure 2). Bivariate analyses on statin use demonstrated the reductions were similar in subgroup analysis among patients with deep vein thrombosis (OR 0.39, 95% CI, 0.2-0.76) and pulmonary embolism (OR 0.29, 95% CI, 0.08-0.97).

Table 4. Bivariate Analysis of Factors Related to Venous Thromboembolism and Occurrence of Venous Thromboembolism

		OR	95% CI
	Diabetes	0.94	0.61-1.45
	Hypertension	0.82	0.56-1.22
	Immobilization	2.52	1.58-4.02
	Metastatic disease	2.44	1.64-3.64
	Current chemotherapy	2.08	1.34-3.22
	Smoking	0.62	0.40-0.98
	Aspirin use	0.57	0.34-0.96
	Statin use	0.33	0.19-0.58

OR=odds ratio; CI=confidence interval.

View Large Image Download to PowerPoint

Figure 2. Forest plot of OR with 95% CI for factors related to venous thromboembolism in the bivariate analysis.

A logistic regression analysis included factors that were statistically significant in the bivariate analyses and have been shown to be related to venous thromboembolism (smoking, documented metastatic disease, current use of chemotherapy, immobilization, use of aspirin, and use of statin), along with statin use. Statin use continued to be related to a decreased odds of developing a venous thromboembolism, similar to the bivariate relationship between statin use and venous thromboembolism (OR 0.33, 95% CI, 0.19-0.59; P<.001). On the other hand, smoking and use of aspirin were no longer statistically significantly related to decreased odds of developing a venous thromboembolism (OR 0.63, 95% CI, 0.39-1.01 and OR 0.69, 95% CI, 0.39-1.2, respectively) (Table 5 and Figure 3).

Table 5. Logistic Regression Analysis of Factors Related to Venous Thromboembolism and Occurrence of Venous Thromboembolism

	Factors	OR	95% CI	P Value
	Immobilization	2.88	1.74-4.75	<.001
	Metastatic disease	2.07	1.34-3.18	.001
	Current chemotherapy	1.77	1.10-2.86	.019
	Smoking	0.63	0.39-1.01	.055
	Aspirin use	0.69	0.40-1.20	.188
	Statin use	0.33	0.18-0.59	<.001

OR=odds ratio; CI=confidence interval. P<.05.

View Large Image Download to PowerPoint

Figure 3. Forest plot of OR with 95% CI for factors related to venous thromboembolism in the logistic regression analysis.

Discussion 

We examined the relationship between factors related to venous thromboembolism and the comorbidities with the occurrence of venous thromboembolism and statin use in hospitalized patients with cancer. In this study, immobilization, documented metastatic disease, and active chemotherapy were all related to increased odds of developing venous thromboembolism. These are well-established risk factors in this population, and this finding is consistent with other recent studies.21, 22

In regard to aspirin use, a meta-analysis on antiplatelet therapy in cardiovascular outcomes showed a significantly reduced risk in the incidence of fatal or nonfatal pulmonary embolism with an odds reduction of 25%.23 However, a recent randomized trial of healthy women demonstrated no clear benefit of aspirin in terms of long-term prevention of venous thromboembolism.24 Our findings are consistent with this result regarding aspirin use.

Smoking seemed to be a protective factor for venous thromboembolism by bivariate analysis. However, smoking no longer statistically reduced the risk of venous thromboembolism in the logistic regression. It seems that patients who did not smoke during the study period were more likely to be immobilized, which was related to an increased risk of developing venous thromboembolism. The inclusion of immobilization in the logistic regression equation might have minimized the effect of smoking.

Our study provides preliminary evidence that statin use might confer protection against the risk of developing venous thromboembolism. There have been many preclinical studies on the effect of statin use and anti-inflammatory effects,25 and evidence that their use might alter vascular endothelial function.26 These studies suggest a pleiotropic effect of statins. On the basis of these theories, statin use might potentially decrease endothelial injury that is prone to propagate thrombus formation. In addition, there is evidence that patients with high levels of apolipoprotein A1 and high-density lipoprotein cholesterol have a decreased risk of recurrent venous thromboembolism.27 These theories might explain the protective effect statin use confers on patients with cancer, who have a high risk of developing venous thromboembolism.

Lastly, statins have a satisfactory safety profile with relatively few adverse effects. The elevation of hepatic transaminase is a common side effect of statin use, but this is often dose-related. For atorvastatin, the rate was 0.2% with 10 mg and 2.3% with 80 mg.28 Warfarin remains the most common preventative medication for venous thromboembolism, but warfarin has multiple concerns and limitations. The therapeutic range needs to be closely monitored, and the case-fatality rate of major bleeding was 9.1% in patients who received anticoagulant therapy for more than 3 months.29 Anticoagulation therapy is contraindicated in patients with active intracranial bleeding30 and needs to be monitored closely in patients with metastatic lesions in the brain, leaving statins as a possible alternative medication for this group of patients.

Limitations 

This study has several limitations. It is retrospective, and there are incomplete records on serum lipid levels and serum markers of inflammation, which could help explain the effect of statins on the occurrence of venous thromboembolism. Moreover, the dosage of each statin was not categorized to distinguish the effect of statins because there might be a dose-relationship response.

Conclusions 

This study provides preliminary data that patients with solid organ tumors receiving statins might have a lower occurrence of developing a venous thromboembolism. A prospective, randomized, placebo-controlled trial would provide further support and stronger evidence for this finding.