Can we prevent venous thrombosis with statins: an epidemiologic review into mechanism and clinical utility

ABSTRACT

Introduction: Statins may be causally associated with a decreased risk of venous thrombosis. If so, this could be a substantive breakthrough since statins do not increase the risk of bleeding and could therefore be used as a safer antithrombotic drug. However, scepticism exists on the observed reduction of venous thrombosis by statins, as it may have been confounded by healthy user effects or other biases.

Areas covered: The main focus of this review will be the biases that may have arisen in clinical studies that investigated the relationship between statin use and risk of venous thrombosis. We also discuss the suggested causal association from a pathophysiological perspective. Furthermore, we integrate the knowledge from clinical and pathophysiological studies into a proposal for new study designs that are needed to sufficiently answer the question whether we can, and should, prevent recurrent venous thrombosis with statins.

Expert commentary: A drug to prevent recurrent venous thrombosis in patients at risk of bleeding that does not induce bleeding and in which the number needed to treat for the prevention of venous thrombosis is sufficiently high, is a remedy that we should continue to look for, and for which statin therapy might be a suitable candidate.

KEYWORDS:

• Venous thrombosis
• statins
• epidemiology
• pathophysiology
• prevention

1. Introduction

Venous thrombosis (deep vein thrombosis or pulmonary embolism) is a common and potentially lethal disease that occurs each year in about 1 to 2/1000 people [1]. The condition can be prevented and treated with anticoagulants, but as a side effect, bleeding often occurs [2]. Currently, the duration of treatment of venous thrombosis with anticoagulants depends on whether the event was provoked or not [3]. Most provoking risk factors, such as surgery, immobilization, and use of oral contraceptives, are of a transient nature. Presence of such a risk factor temporarily increases the ‘thrombotic potential’ of an individual, and hence, the risk decreases once the risk factor is gone. This explains, for example, why recurrence risk is low (<1%/year) in patients who developed their first event after surgery [4]. Patients with provoking risk factors are usually treated with anticoagulants for 3–6 months only, while patients with unprovoked thrombosis are prescribed anticoagulant treatment for a longer period [3]. This extended treatment should be seen as prevention of a recurrence, in which decision is based on its high incidence in patients with unprovoked events (30% within 5 years after the 3–6 months of oral anticoagulation) [5]. Only 40–50% of all thrombosis patients can be classified as patients with a first provoked event, which leads to a dilemma in the other 50–60%: discontinuing treatment may lead to a new thrombotic event, while continuing oral anticoagulant treatment is accompanied with an yearly 1–3% risk of major bleeding [2,3]. Therefore, novel therapeutic strategies to prevent venous thrombosis that are not associated with bleeding complications are urgently needed.

In this review, we will discuss whether statins are causally associated with a decreased risk of venous thrombosis. If true, this could be a substantive breakthrough since statins are known not to cause bleeding. We will summarize clinical research that studied whether statins exert beneficial effects in preventing venous thrombosis. Next, we will discuss possible mechanisms from a pathophysiological perspective. Finally, we will integrate the knowledge obtained from clinical and pathophysiological studies into a proposal that is needed to sufficiently answer the question if we can prevent recurrent venous thrombosis with statin therapy.

2. Clinical studies

2.1. Initial findings

In 2000, Grady and colleagues were the first to report that statin use was associated with a 50% reduced risk for development of venous thrombosis in postmenopausal women starting estrogen and progestin therapy [6]. Since then, many other studies have been published on the association between statin use and decreased risk of venous thrombosis. For instance, a meta-analysis of seven observational studies revealed that statin use was associated with a significantly lower risk of venous thrombosis compared to non-statin use (odds ratio, 0.62, 95% confidence interval [CI], 0.45–0.86) [7]. However, statin use is associated with several preventive effects in observational studies: it is associated not only with lower risks of venous thrombosis, but also with lower risks of arrhythmia, multiple sclerosis, Alzheimer’s dementia, infections, AIDS, cancer mortality, and even motor vehicle accidents [8–10]. Because these effects do not seem to be due to lower lipid levels, this raises the suspicion that at least some of the observed associations are due to noncausal mechanisms, such as bias.

2.2. Bias

Bias can be defined as a process at any stage of causal inference which tends to produce results or conclusions that differ systematically from the truth. The landmark article by Sackett in 1979 listed 35 biases [11], of which we will discuss three types that could explain the observation that statins are associated with a decreased risk of venous thrombosis. The first lies in what has been called the ‘healthy-user’ effect, i.e. that statins are prescribed preferentially to individuals with a favorable risk profile or that the healthiest users are analyzed in some observational studies [12]. However, for venous thrombosis, it is unlikely that this healthy-user effect fully explains the positive association with statin use, as (high risk of) arterial cardiovascular disease is an indication for statin use, of which some risk factors (age, male sex, obesity, and smoking) are shared risk factors for both conditions [13]. Thus, participants in observational studies who use statins should have a less favorable cardiovascular risk factor profile than nonusers and are therefore at higher (and not lower) baseline risk for venous thrombosis.

However, other types of bias may have contributed to the observed lower risk of venous thrombosis while using a statin, for instance, in studies that included individuals who had been using statins for some time prior to study entry [7]. Such ‘prevalent users’ can introduce two types of bias: (1) underascertainment of events that occur early after starting treatment (survivor bias) and (2) the inability to control for those who do or do not adhere to statin treatment (adherence bias) [14,15].

In terms of survivor bias, for statins, there is indirect evidence that the risk for venous thrombosis is increased in the first months of statin treatment, since an indication for statin therapy is a recently experienced arterial cardiovascular event. Because it has been reported that patients with acute arterial cardiovascular disease are at increased risk of subsequent venous thrombosis and death for a short time period [16,17], underascertainment of venous thrombotic events can result from early attrition of patients on statins who are most susceptible but may have died or be too sick to be enrolled in an observational study.

In terms of adherence bias, prevalent users in a study, by definition, use a statin at time of inclusion, while those who had an indication for statin treatment yet failed to continue with their treatment are abraded as non-statin users. Adherence to a drug is a marker for a constellation of unmeasured factors and likely associated with better outcome, independent of the drug use itself. This is true for all drugs, including statin use and even for placebo use. For instance, several randomized controlled clinical trials in which patients who were adherent to placebo showed 30–60% reduced risks of death from cardiovascular disease compared with non-adherent placebo users, in which magnitude of the association was not materially affected by adjustment for several potential confounders [18–20].

For these two reasons, it is important to take prevalent users into account when studying effects of statin treatment on the risk of venous thrombosis in observational studies. For this, one can use a so-called ‘new-user design’ [14]. Such a design begins by identifying all individuals in a predefined population (both in terms of people and time) who for the first time start a course of treatment with a statin. Study follow-up for end points begins at precisely the same time as initiation of statin therapy or t = 0. Data for all patient characteristics are obtained at a time just before t = 0. Observational studies can be performed by initially assembling a cohort consisting of only new users and an appropriate comparison group or by identifying new users and the comparison group from an existing cohort. This definition is similar to the way in which data are analyzed in a clinical trial, where t = 0 is the time of randomization (usually just before treatment begins), except of course that treatment is not assigned by randomization. A new-user design differs from most observational studies in that it excludes prevalent users. For this matter, it is interesting to note that in aforementioned meta-analysis [9], only one of the seven mentioned studies included statin initiators in their study (new-user design) and reported an odds ratio of 1.02 (95% CI, 0.88–1.18) [7,21], which is in contrast to the overall odds ratio of 0.62 (95% CI, 0.45–0.86). This raises the suspicion that the observed association between statin use and a decreased risk of venous thrombosis suffered from prevalent users in these studies leading to bias. However, other studies in which prevalent user bias was excluded by design have been published, including results from a trial and a meta-analysis of trials [22,23], which we will discuss below.

2.3. Class effect?

Side effects of drugs are not necessarily class effects, particularly when the mechanism of the side effect differs from the primary mechanism of the drug. It is known that the mechanism of statins varies between the types of statins that are currently on the market today, showing different reducing effects on low-density lipoprotein, atherosclerosis, and inflammation. This reduction is the least strong in pravastatin users, followed by simvastatin users and atorvastatin users and is strongest in rosuvastatin users [24,25].

There are some studies that suggest that dyslipidemia, inflammation, or atherosclerosis, i.e. determinants for arterial cardiovascular disease, also increase the risk of venous thrombosis [26–28]. Therefore, an analysis done by type of statin to venous thrombosis risk seems sensible as, in case of a causal association mediated through dyslipidemia, inflammation, or atherosclerosis, the effect on venous thrombosis risk should be strongest in rosuvastatin and weakest in pravastatin users. As summarized in Table 1, in a cohort study of nearly 2 million individuals from the United Kingdom, in which a new-user design was used, the authors showed that rosuvastatin use was associated with the strongest (approximately 40%) reduced risk of venous thrombosis [29]. These results closely resemble the results from randomized controlled trials [22,23]. For the occurrence of venous thrombosis, a predefined analysis of a randomized clinical trial in which apparently healthy individuals were randomized to rosuvastatin or placebo (JUPITER trial) showed a 40% risk reduction when using rosuvastatin compared with placebo (Table 1) [22]. In the absence of other randomized trials with venous thrombosis as the primary end point, Rahimi and colleagues presented a pooled analysis of 29 randomized statin studies in which venous thrombotic events were reported as serious adverse events [23]. They failed to confirm a risk reduction of venous thrombosis by statin treatment. However, the authors found that individuals who were randomized to rosuvastatin still had an approximately 40% reduced risk of venous thrombosis (hazard ratio, 0.65; 95% CI, 0.33–1.28) (Table 1). Albeit CIs were wide in the study from Rahimi et al., results from Table 1 suggest a dose–response relation where the statin that is most related with halting/regression of atherosclerosis, dyslipidemia, and inflammation (i.e. rosuvastatin) also provides the largest risk reductions for the occurrence of venous thrombosis.

Table 1. Effect of statin therapy on venous thrombosis by type of statin.

	Observational study [30]		Clinical trials [24,25]
	Hazard ratio (95% CI)		Hazard ratio (95% CI)
	Women	Men	JUPITER trial	Rahimi et al.^a
No statin	1 (reference)	1 (reference)	1 (reference)	1 (reference
Pravastatin	1.05 (0.80–1.38)	1.00 (0.76–1.31)		1.17 (0.81–1.69)
Simvastatin	0.91 (0.83–0.99)	0.88 (0.80–0.97)		0.94 (0.71–1.25)
Atorvastatin	0.86 (0.76–0.99)	0.86 (0.75–0.97)		0.82 (0.56–1.19)
Rosuvastatin	0.61 (0.36–1.03)	0.53 (0.29–0.95)	0.57 (0.37–0.86)	0.65 (0.33–1.28)
Any statin	Unavailable	Unavailable		0.93 (0.82–1.07)

^aExcluding results from the JUPITER trial.

CI: confidence interval; JUPITER: Justification for the Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin.

3. Pathophysiology

3.1. Statins and (early) atherosclerosis

In 2003, the hypothesis was sparked that atherosclerosis leads to venous thrombosis [31]. As in this study atherosclerosis measurements were performed after venous thrombosis occurred, its temporal relation and causality were not entirely clear [30]. Currently, there is little evidence available that venous thrombosis and atherosclerosis are causally associated [13]. However, there is some biological evidence that may give credence to the causality of this association since the hemostatic system seems to be able to accelerate atherosclerosis [30]. This was demonstrated in mouse studies with hypercoagulable and diminished coagulation phenotypes on an atherosclerotic background, where diminished coagulation provided protection against atherosclerosis development, whereas hypercoagulable mice developed more severe atherosclerosis [32]. In human histological studies, it has been shown that a procoagulant state is more abundantly present in early-stage atherosclerotic lesions than in advanced atherosclerotic lesions [33–35]. Why coagulation factors are more abundantly present within early atherosclerotic vessels than in advanced atherosclerosis is as yet unknown, but may be attributable to primary protective mechanisms against vascular injury [36]. With the advent of in vivo carotid magnetic resonance imaging (MRI) screening, one can now distinguish early from advanced atherosclerosis [37,38]. It is therefore possible to perform clinical studies to quantify as to whether both early and/or advanced atherosclerosis increases the risk of venous thrombosis. Such studies could clarify why all of the statins that are currently available, the most potent anti-atherosclerotic ones, are associated with a lower risk of venous thrombosis [22,23,29]. As far as we know, such studies have not been conducted yet.

3.2. Statins and dyslipidemia

As lipid levels can be modulated by lifestyle intervention and statin therapy [39], the potential association between lipids and venous thrombosis and its underlying pathophysiology is a relevant issue worth pursuing. However, whether lipid levels themselves are associated with venous thrombosis is controversial due to different results among epidemiological studies. A previous meta-analysis demonstrated that mean levels of total cholesterol were higher and HDL cholesterol levels were lower in venous thrombosis patients than in controls [26]. However, the majority of the reports on lipids and venous thrombosis in this meta-analysis were small case-control studies, and individually controlling for several confounders was not possible [26]. Moreover, there was severe heterogeneity between studies meaning that there was a large variation in study outcomes between studies, of which the larger ones found no association with lipid levels and venous thrombosis. Several large population-based cohort studies have been published since the aforementioned meta-analysis, and these additional studies found no association between dyslipidemia and venous thrombosis after controlling for confounding factors and competing risk [40–42]. In addition, non-statin lipid-lowering drugs (i.e. fibrates) are not associated with a reduced venous thrombosis risk [43]. Therefore, we consider it unlikely that statins decrease venous thrombosis risk by their lipid-lowering activities.

3.3. Statins and inflammation

Several lines of evidence, ranging from experimental models to population-based studies, support the notion that inflammation is a driver of atherosclerosis [44]. From an epidemiological perspective, numerous studies have shown that the inflammatory marker high sensitivity c-reactive protein (hs-CRP) is associated not only with atherosclerosis, but also with an increased risk of venous thrombosis [45,46] and with higher levels of procoagulant factor VIII [47]. However, a Mendelian randomization study convincingly showed that hs-CRP levels are not a cause of venous thrombosis [47]. Still, statins, initially manufactured to target dyslipidemia and slowdown atherosclerosis, showed that they also have anti-inflammatory properties [48]. Since atherosclerosis can produce both an inflammatory and a procoagulant response [30,44], reduction of inflammation by statins could be a driving force behind the reduced venous thrombosis risk that has been observed in statin users. Interestingly, one study of 26 patients who had venous thrombosis found that a 3-day administration of atorvastatin reduced inflammation as evidenced by reduced interleukin (IL)-6, IL-8, and soluble P-selectin, together with increased anti-inflammatory IL-10, without any significant effect on hs-CRP [49]. Because of the short time interval, this study suggests potential benefits from statin administration with regard to reduced venous thrombosis risk that is in part driven by an immediate anti-inflammatory (i.e. not related with atherosclerosis) effect.

3.4. Statins and platelet activation

Another mechanism through which statins may decrease the risk of venous thrombosis is by inhibiting platelet activation and consequently aggregation [50]. Animal models have shown that platelet activation plays a key role in the initiation of thrombus formation in deep vein thrombosis [51,52]. Furthermore, in vitro studies indicate that statins inhibit platelet activation via several lipid-independent mechanisms including the inhibition of thromboxane A2 (TxA2) formation [53]. Since enhanced platelet aggregation has been reported in patients with venous thrombosis [54], we recently decided to investigate if there is an effect of rosuvastatin on TxA2-mediated platelet activation in individuals with a history of venous thrombosis in the STAtins Reduce Thrombophilia trial (START) trial [55]. We randomized 25 individuals to rosuvastatin 20 mg daily for 28 days and 25 individuals to no statin and observed no effect of TxA2-mediated platelet activation in rosuvastatin users. These findings show that it is unlikely that the association of a decreased risk of venous thrombosis in rosuvastatin users is explained by decreased TxA2-mediated platelet activation. However, a limitation of this study was that only one assay was used to evaluate platelet function [56]. Therefore, we cannot exclude potential other antiplatelet effects of rosuvastatin as can be measured using different platelet function tests.

3.5. Statins and coagulation

Colli et al. were the first in 1997 to report that statins interfere with activation of the clotting system and the coagulation cascade [57]. In that study, the authors observed that tissue factor was suppressed in macrophages that were incubated for 20–24 h with simvastatin. Many reports followed [58,59], for example, by Undas and colleagues who showed that once-daily use of simvastatin 40 mg given for 3 days in 14 healthy volunteers resulted in a reduction of thrombin formation [60], that was of a similar magnitude to that observed after 90 days of statin therapy [61]. Another study found evidence for rapid alterations in fibrin clot structure/function induced by statins in venous thrombosis patients treated with atorvastatin 40 mg once daily for 3 days [62]. Based on these reports, the authors suggested that statin use may decrease the risk of venous thrombosis by downsizing coagulation activation. However, these findings should be interpreted with caution as not all studies consistently reported a favorable outcome of statin treatment on coagulation factor levels. For example, in a randomized study from Dangas et al. (n = 93), an increase in fibrinogen level was observed when individuals were exposed to pravastatin compared with placebo [63]. Recently published findings from randomized controlled trials of statin therapy suggest some effect of lowering levels of von Willebrand factor and D-dimer [64,65]. However, funnel plot analyses showed that these positive results might be due to publication bias. Furthermore, only one trial (n = 60) reported on the potential effects of rosuvastatin to the hemostatic system [66]. Observational studies that found a possible relation between statin use and a decreased level of procoagulant factors [67] could have been hampered by methodological issues such as survivor bias and adherence bias as they included prevalent users [14]. In addition, none of these studies may be generalizable to venous thrombosis patients as they have generally been conducted in patients with hyperlipidemia, diabetes, or other disease states unrelated with venous thrombosis.

In an observational study with an active-comparator design, the drug of interest is compared with another agent commonly used for the same indication, rather than with no treatment (a ‘nonuser’ group) [68]. This principle helps to ensure that treatment groups have similar treatment indications, attenuating both measured and unmeasured differences in patient characteristics. Studies that used such a design on the issue as to whether coagulation factor levels are influenced by statins have, as far as we know, not been reported previously. For the current review, we explored such a study design in the control individuals of the Multiple Environmental and Genetic Assessment of risk factors for venous thrombosis (MEGA) study of which results are shown in Table 2. We analyzed if there was a difference in the median level of several coagulation factors in individuals who were treated with lipid-lowering drugs, i.e. either statins or fibrate, at time of blood sampling. Although from Table 2 it appears that of the lipid-lowering therapies that were prescribed, statin users had a less hypercoagulable profile than fibrate users (with rosuvastatin users having the most favorable coagulation profile), we believe that this finding should be interpreted with caution as numbers are small.

Table 2. Median levels of procoagulant factors in fibrate and statin use in control subjects of the MEGA study.

		Statin users
Median level (IQR)	Fibrate users (n = 7)	Fluvastatin (n = 10)	Simvastatin (n = 144)	Atorvastatin (n = 136)	Pravastatin (n = 53)	Rosuvastatin (n = 11)
Procoagulant
Fibrinogen, g/L	3.3 (2.5–4.0)	4.0 (3.3–4.3)	3.5 (3.1–4.0)	3.6 (3.1–4.1)	3.5 (3.2–4.1)	3.6 (3.1–4.0)
Factor II,†,^a IU/dL	111 (103–122)	109 (97–127)	113 (104–122)	110 (101–118)	114 (133–127)	102 (88–124)
Factor VII,^a† IU/dL	139 (110–152)	121 (99–134)	118 (98–138)	124 (107–139)	127 (118–153)	97 (78–129)
Factor VIII:Ag, IU/dL	123 (70–204)	133 (109–205)	130 (106–159)	114 (92–154)	119 (84–145)	101 (86–137)
VWF, IU/dL	111 (58–170)	115 (91–126)	118 (96–159)	111 (87–150)	103 (71–144)	95 (85–137)
Factor IX,^a† IU/dL	114 (103–159)	118 (99–142)	111 (101–125)	109 (100–124)	112 (98–125)	101 (92–109)
Factor X,^a† IU/dL	118 (102–146)	111 (103–132)	119 (104–130)	120 (110–133)	120 (100–133)	111 (98–125)
Factor XI, IU/dL	115 (110–156)	101 (86–131)	105 (90–121)	104 (94–117)	104 (88–110)	93 (86–122)

^aControl subjects on vitamin K antagonists excluded.

IQR: inter quartile range; MEGA: Multiple Environmental and Genetic Assessment of risk factors for venous thrombosis (MEGA) study; VWF: Von Willebrand factor.

A powerful design not suffering from any of the aforementioned limitations to answer the question whether coagulation factors are influenced by statin therapy is a sufficiently powered randomized trial in which the primary outcome is change in coagulation factor level in patients with prior venous thrombosis. We are currently conducting such a trial (START study; www.clinicaltrials.gov; NCT01613794) in which patients, after having received treatment with anticoagulants for venous thrombosis, are randomized to rosuvastatin 20 mg once daily for 1 month to study the potential anticoagulant properties of statins. The study is powered on coagulation factor VIII, as a high factor VIII level is well associated with both first and recurrent venous thrombosis [69–71]. Of note, START will only look at the immediate (1 month) effect of rosuvastatin treatment to coagulation factor levels. Randomized studies that take longer time effects of rosuvastatin therapy to coagulation factor levels into account may be covered by the StAtins for Venous Event Reduction in Patients With Venous Thromboembolism Pilot Study trial (www.clinicaltrials.gov; NCT02679664).

4. Expert commentary

Whether venous thrombosis can be prevented with the use of statins is questionable as the available data are scant, controversial, and likely biased, with few replicated studies. Nevertheless, pathophysiological insights reveal a potential mechanism through anticoagulant and/or anti-inflammatory properties of statins, most notably rosuvastatin, possibly by targeting (early) atherosclerosis. That cheap and accessible anti-atherosclerotic drugs, such as statins, might prevent venous thrombosis without inducing bleeding offers currently not enough ground to start statin treatment in these patients. For this, we need more and better evidence.

5. Five-year view

The following research questions and accompanying designs are essential to answer the question if statins can prevent venous thrombosis.

1. We need to know the pharmacological mechanism of how statins are able to decrease the risk of venous thrombosis. This should be studied in sufficiently powered randomized clinical trials or in active comparator study designs that are externally validated and primarily set up for this reason. We are currently conducting the START trial (www.clinicaltrials.gov; NCT01613794) to analyze the immediate 1-month effects on coagulation of 20 mg rosuvastatin use once daily in 250 randomized patients with prior venous thrombosis. Since rosuvastatin is the statin that consistently showed the strongest association between its intake and a reduced risk of venous thrombosis [22,23,29], we will consider a negative signal from this trial as negative for the whole class of statins. A possible positive signal may not rule in potential beneficial effects on the coagulation profile from the other statins that are currently on the market. Of note, START can only look at the immediate (1 month) effect of rosuvastatin treatment to coagulation factor levels.

2. We need to know if various stages of atherosclerosis can increase the risk of venous thrombosis. If so, the culprit behind the supposed causal association between statin intake and reduced coagulation could be early atherosclerosis. Of the latter, previous research has shown that increased coagulation activity is specifically observed in early atherosclerotic lesions [30,33], but we do not know if early atherosclerosis increases venous thrombosis risk. To study this, we need to perform a clinical study (either case-control or follow-up) in which early and advanced stages of atherosclerosis are viewed accurately in vivo, which is now possible with noninvasive high-resolution MRI [37,38]. External validation of results from a previous autopsy study [33] is needed to endorse that early atherosclerosis is also associated with an ex vivo prothrombotic state/venous thrombosis compared with no atherosclerosis or advanced atherosclerosis, with rigorous adjustments for potential confounding factors.

3. If the hypotheses in abovementioned study proposals turn out to be true, this will not immediately change clinical practice as for this a randomized trial with clinical end points is needed. Large randomized trials are not started in a vacuum, but they require a high prior probability of success, some insight in pathophysiological mechanisms, and need to be performed in specific groups who may benefit the most. The results of the studies as proposed in (1) and (2) will provide sufficient evidence to determine if such a trial should be conducted. Since the question as to whether a therapy is effective is answered by the number needed to treat (NNT) per year [72], the secondary prevention of venous thrombosis after an unprovoked event with statins, where reasonable NNTs may be plausible because of high recurrent venous thrombosis event rates [73–75], could be feasible. It is in this aspect interesting to note that the current American College of Chest Physicians guideline suggests to discontinue anticoagulant treatment in patients with venous thrombosis who are considered to be at high risk of anticoagulation-related bleeding [3]. Although statins are unlikely to be as effective as anticoagulant drugs, they do have the major advantage over anticoagulants that they do not induce bleeding. Therefore, a drug to prevent recurrent venous thrombosis in patients at risk of bleeding that does not induce bleeding and in which the NNT for the prevention of recurrence is sufficiently high is a remedy that we should continue to look for, and for which, statin therapy might be a suitable candidate.

Key issues

• Venous thrombosis is a common and potentially lethal disease which can be prevented and treated with anticoagulants, but as a side effect serious bleeding often occurs

• Therefore, novel therapeutic strategies that are not associated with bleeding complications are urgently needed

• Statins may form a suitable alternative to prevent venous thrombosis as they do not induce bleeding but may reduce the risk of venous thrombosis

• Pathophysiological insights reveal that statins exhibit several vascular protective effects, including anti-inflammatory and antithrombotic properties, that are not related to changes in lipid profile

• However, whether venous thrombosis can be prevented with the use of statins is questionable as the available data are scant, controversial and likely biased, with few replicated studies

• This review integrates knowledge from clinical and pathophysiological studies into new study designs that are needed to answer the question if we can prevent recurrent venous thrombosis with statin therapy

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

This paper was not funded.