10945_Statin Therapy Associated With Improved Thrombus Resolution In Patients With Deep Vein Thrombosis

Yale University
Yale University
EliScholar – A Digital Platform for Scholarly Publishing at Yale
EliScholar – A Digital Platform for Scholarly Publishing at Yale
Yale Medicine Thesis Digital Library
School of Medicine
January 2020
Statin Therapy Associated With Improved Thrombus Resolution In
Statin Therapy Associated With Improved Thrombus Resolution In
Patients With Deep Vein Thrombosis
Patients With Deep Vein Thrombosis
Charles Hsu
Follow this and additional works at: https://elischolar.library.yale.edu/ymtdl
Recommended Citation
Recommended Citation
Hsu, Charles, “Statin Therapy Associated With Improved Thrombus Resolution In Patients With Deep Vein
Thrombosis” (2020). Yale Medicine Thesis Digital Library. 3911.
https://elischolar.library.yale.edu/ymtdl/3911
This Open Access Thesis is brought to you for free and open access by the School of Medicine at EliScholar – A
Digital Platform for Scholarly Publishing at Yale. It has been accepted for inclusion in Yale Medicine Thesis Digital
Library by an authorized administrator of EliScholar – A Digital Platform for Scholarly Publishing at Yale. For more
information, please contact elischolar@yale.edu.

Statin Therapy Associated with Improved Thrombus Resolution
in Patients with Deep Vein Thrombosis

A Thesis Submitted to the
Yale University School of Medicine
in Partial Fulfillment of the Requirements
for the Degree of Doctor in Medicine

By
Charles Hsu
2020

1
Abstract
Objectives: Statin therapy has been associated with a decreased incidence of venous
thromboembolism (VTE) in clinical trials and enhanced thrombus resolution in animal
models. The effect of statins on thrombus resolution has not been reported clinically. This
study investigates the association of statins with thrombus resolution or improvement in
patients with deep vein thrombosis (DVT).

Methods: A retrospective study of the electronic medical records of consecutive adult
patients presenting with lower extremity DVT was performed. Patients were divided into
two groups based on statin therapy (statin group) or lack thereof (non-statin group). The
two groups were compared with respect to demographics, comorbidities, and risk factors
for VTE. Initial as well as all subsequent ultrasound reports were reviewed for each
patient to determine extent of DVT and subsequent change in thrombus characteristics.
Long-term outcomes examined were thrombus improvement or resolution on follow up
ultrasound, VTE recurrence, mortality. Multivariable analysis was used to determine
independent predictors of thrombus resolution or improvement, VTE recurrence, and
mortality.

Results: A total of 818 patients with DVT were identified [statin group: n = 279 (34%),
non-statin group: n = 539 (66%)]. The patients in the statin group were significantly older
(P < .001). Patients on statin were more likely to have risk factors for and manifestations of atherosclerosis and to be on antiplatelet therapy (P < .001) while those in the non-statin group were more likely to have a hypercoagulable disorder (P = .009) or prior DVT (P 2 = .033). There was no significant difference in provoked DVT, extent of DVT, or association with PE (pulmonary embolus), but patients on statin were more likely to have high-risk PE (P = .046). There was no difference in patients receiving anticoagulation, type and duration of anticoagulation, inferior vena cava filter placement, or treatment with lytic therapy. There was no difference in thrombus resolution, mortality, or recurrence of DVT, PE, or VTE between the groups. On multivariable analysis, age, proximal DVT, CAD, and cancer were associated with higher mortality while anticoagulation with warfarin and DOACs and antiplatelet therapy were associated with lower mortality. Statin therapy, antiplatelet therapy and younger age were associated with thrombus resolution or improvement. Conclusions: Statin therapy is associated with greater thrombus resolution or improvement in patients with DVT. However, statin therapy in this study was not associated with different clinical outcomes of VTE recurrence or mortality. 3 Acknowledgements: This work was performed under the guidance of Cassius Iyad Ochoa Chaar, MD, MS, Associate Professor of Surgery, the Division of Vascular Surgery, Department of Surgery, Yale University School of Medicine. Other contributors to this work include Anand Brahmandam, MD, Kirstyn E Brownson, MD, Nancy Huynh, MD, Jesse Reynolds, MS, Alfred I Lee, MD, PhD, and Wassim H Fares, MD, MSc. Contributions are described in detail in the Methods section. This work was made possible by the grant from the National Heart, Lung and Blood Institute of the National Institutes of Health. The content is solely the responsibility of the author and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute or the National Institutes of Health under award Number T35HL007649. 4 Table of Contents Page Introduction 5 Aim 12 Methods 13 Results 18 Discussion 22 Conclusion 31 References 32 Tables 39 5 Introduction: Venous thromboembolism (VTE) is a common condition affecting hundreds of thousands of patients every year, including those with deep vein thrombosis (DVT) and pulmonary embolism (PE), a potentially life-threatening condition.1 Patients with VTE are often managed using anticoagulation agents, the medications of choice for the prevention and treatment of VTE, however there remains an unmet need for those who cannot tolerate their side effects or may be inadequately treated. Studies suggest that statins, also known as 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase inhibitors and most commonly known for their cholesterol-lowering effects, may also play a role in reducing the risk of initial or recurrent DVT.2,3 In addition, population studies of VTE patients on statins have shown that longer duration of statin use is also associated with greater reduction of recurrent VTE.4 The precise mechanisms of this observed antithrombotic effect have been explored in a variety of preclinical experiments. In animals, atorvastatin and rosuvastatin have been shown to reduce venous thrombus burden and DVT-induced vein wall scarring, while simvastatin has been found to promote thrombus resolution in a leporine posterior vena cava thrombus model.5,6 Clinically, residual vein thrombosis has been shown to be a risk factor for VTE recurrence.7 Thus, it is plausible that the effects of statins on VTE recurrence are potentially mediated by enhancing thrombus resolution. However, the effect of statins on thrombus remodeling has not been studied clinically. The following sections review both the nonclinical and clinical literature supporting the above concepts to establish the context and rationale for the work performed in this thesis. 6 Anti-thrombotic effects of statins on the coagulation cascade Statins are most commonly known for their efficacy in the management of hyperlipidemia and its associated sequelae (i.e. coronary artery disease), and have also shown to significantly reduce the risk of major cardiovascular events.8 However, statins have also been shown to provide cardiovascular benefit via a number of “pleiotropic” effects independent from cholesterol lowering, notably including anti-thrombotic properties at multiple levels of the coagulation cascade.9 The lipophilic statins simvastatin and fluvastatin (but not the hydrophilic pravastatin) decrease tissue factor (TF) expression in macrophages via nuclear factor κB (NF-κB) inhibition in vitro, an effect which was not reversed by the addition of cholesterol, suggesting that statin’s effect on TF occurs independently of intracellular cholesterol lowering. Instead, a different downstream product, geranylgeranyl pyrophosphate (GGPP) is thought to play a crucial role in the regulation of TF expression.10 As TF expression has been shown to be upregulated in lipid-laden macrophages at the core of atherosclerotic plaques and is thought to promote intravascular thrombosis, the down-regulating effect of statins on TF would explain another cholesterol-independent aspect of statins’ cardioprotective properties.11 Indeed, a study in humans found that carotid artery plaques removed from patients treated with atorvastatin for 4-6 months had significantly lower TF antigen levels and activity as compared with those on placebo.12 Further downstream in the coagulation cascade is thrombin, the key to activating thrombus formation and the common target of several marketed anticoagulant therapies. Statins, particularly atorvastatin, simvastatin, and pravastatin, have been shown in both 7 experimental and clinical studies to decrease thrombin formation, however measurements of thrombin markers such as prothrombin fragments 1.2 (F1.2) and thrombin- antithrombin complex (TAT) have yielded less consistent findings, ranging between mild to neutral effects. This effect is particularly notable in patients with hypercholesterolemia, where even 3-days of simvastatin (40mg/d) treatment in fourteen men with LDL cholesterol levels >130 mg/dL on low-dose aspirin (75mg/d) resulted in a
significantly reduced level of prothrombin activation at sites of microvascular injury,
accompanied by delayed Factor Va generation and accelerated activated Protein C-
mediated Factor Va inactivation, on top of the anticoagulant effects of aspirin alone.13-15

To a lesser degree of scientific concordance, statins have also been found to exert anti-
thrombotic effects on the coagulation cascade via decreased fibrinogen cleavage,
decreased Factor XIII activation, increased thrombomodulin expression, and increased
Factor Va inactivation.16-18 Other pleiotropic effects of statins include improved
endothelial function via increased nitric oxide bioavailability, inflammation suppression,
atherosclerotic plaque stabilization, immune-modulation, inhibition of cardiac
hypertrophy, and reduced smooth muscle cell proliferation.16 These mechanisms, in
concert with each other, are likely to underpin the anti-thrombotic and vascular protective
effects observed in animal and clinical studies.

Statins and the risk of venous thromboembolism
The clinical relevance of the antithrombotic effects of statins has been investigated by
many studies. In 2000, the Heart and Estrogen/Progesterone Replacement Study (HERS)
8
found in a secondary analysis of the impact of hormonal therapy on the risk of VTE that
use of statins (but not any other lipid-lowering drug) in a postmenopausal population was
associated with a decreased risk of VTE (relative hazard 0.5 [0.2 – 0.9]).19 In 2001, Ray
et al. showed through a large Canadian retrospective cohort study that any dose of statin
use in individuals greater than 65 years old was associated with a 22% lowered incidence
of venous thromboembolisms.20 This association has been supported by a number of
case-control, cohort, or observational studies, including studies finding statin-associated
VTE risk reduction in patients with underlying hypercoagulable states such as cancer and
nephrotic syndrome.21-23 One study by Ramcharan et al. found that different statin types
and treatment durations were all associated with a decreased risk of VTE (OR 0.45 [0.36
– 0.56]).23 However, not all studies have found a positive association between statins and
reduced VTE risk, including a population-based retrospective follow-up nested case-
control analysis of the United Kingdom’s General Practice Research Database and a
prospective open cohort study in England and Wales.24,25 Considering the divergent
findings in these various retrospective studies, perhaps the single most important trial is
then the randomized, double-blinded, placebo-controlled Justification for the Use of
statins in Prevention: an Interventional Trial Evaluating Rosuvastatin (JUPITER) study
published in 2009. In the JUPITER study, the trialists found that 20 mg/d of rosuvastatin
reduced the risk of VTE in asymptomatic patients (men over 50 years old and women
over 60 years old) with no evidence of cardiovascular disease, LDL-cholesterol <130 mg/dL, and CRP >2 mg/L (at first visit) over the course of 1.9-years of follow-up. The
study found significant reductions in DVT or PE (34 events in the treatment arm vs 60
events in the placebo arm, HR 0.57 [0.37–0.86]), with the largest reduction in isolated
9
DVT (55%), however failed to reduce the occurrence of PE (OR: 0.77 [ 0.41–1.46]) or
VTE death (OR: 0.50 [0.20–1.24]).2 A meta-analysis of 4 cohort and 4 case-control
studies published in 2011 by Pai et al. sought to expand investigation beyond the healthy
older adults studied in JUPITER to include a more heterogeneous (younger or comorbid)
populations. Pai et al. found that while statin use was significantly associated with a
lower rate of VTE (OR 0.67; 95% CI 0.53 – 0.84), its association with lower rate of DVT
(OR 0.53; 95% CI 0.22 – 1.29) did not meet significance and was likely attributable to a
small sample size. The body of evidence taken together suggest that while statins’ effect
on reducing VTE may be variably observed based on population and underlying medical
comorbidities, the possibility of a clinically meaningful association deserves further
study.

Residual thrombus in DVT and the post-thrombotic syndrome
There is a myriad of risk factors associated with the recurrence of VTE, and one of
interest is residual venous thrombus. Many studies have attempted to understand the
individual risk of VTE recurrence to better align treatment duration with patient-specific
benefit-risk profiles. In a 2011 systematic review, Tan et al. found that residual
thrombosis was positively associated with VTE recurrence, suggesting that assessing
residual venous obstruction may be useful in individual risk assessment.26 However, this
finding did not seem to translate to clinical management, as the contemporaneously
conducted REVERSE cohort study by L. E. Gal et al. did not find a significantly higher
risk of recurrent VTE in patients who were found to have residual thrombus at the time of
completing 5-7 months of anticoagulant therapy.27 While this finding does not support
10
the assessment for residual thrombus in the clinical determination of individual treatment
duration, the presence of a residual thrombus remains pathologically significant given the
chronic sequelae of venous hypertension and vein wall scarring which characterizes post-
thrombotic syndrome (PTS).28,29 In PTS, typical symptoms include pain, swelling, and
cramping of the leg, generally worsened when standing, and in severe cases can lead to
venous ulceration. Despite optimal anticoagulation by conventional therapies, many
patients with iliofemoral venous thrombosis continue to develop PTS, highlighting an
unmet medical need, particularly around thrombus resolution. While the prevention of
PTS is rooted in DVT prevention and the active management of incident DVT’s, cases of
extensive DVT can be treated with open or endovascular thrombus removal (e.g.
catheter-directed thrombolysis).28,30 In the 2016 CaVenT trial, Haig et al. showed that
patients randomized to catheter-directed thrombolysis by alteplase within 21 days from
symptom onset plus standard therapy (anticoagulation therapy and compression
stockings) were less likely to develop PTS compared to those on standard therapy alone
at 5 years follow-up.31 This supports the notion that thrombus resolution may be
associated with improved outcomes and suggests that in the absence of intervention
(which may not be universally accessible) there may be a role for medications that
promote thrombus resolution beyond currently available anticoagulants.

Statins and thrombus resolution
In addition to its lipid-lowering and antithrombotic effects, statins have also been found
to improve the resolution of venous thrombi. In a murine study published in 2015,
Kessinger et al. demonstrated a 25% thrombus burden reduction in mice with stasis-
11
induced (IVC ligation) venous thrombi treated with atorvastatin or rosuvastatin, which
was accompanied by a significant reduction in platelet aggregation and clot stability, as
well as anti-inflammatory effects most notable early after thrombus formation, and a 50%
reduction in thrombus-induced vein wall scarring.5 A similar model in rabbits was later
studied by Feng et al., finding that simvastatin increased thrombus resolution, likely
through anti-inflammatory effects independent of its lipid-lowering properties.6 The
authors also noticed that simvastatin affected thrombus size sooner than low molecular
weight heparin (3-day vs 14-day), with effects equalizing in the 14-day arm, likely
explained by their different mechanisms of action. These preclinical models suggest that
statins may be an effective treatment in patients with VTE who cannot otherwise tolerate
currently available options for anticoagulation. While there is evidence from a single case
report of aggressive combination statin and antiplatelet therapy leading to rapid
regression of an unstable mobile thrombus aortic atheroma, no other clinical studies have
been reported exploring the effect of statins on thrombus resolution in humans.32

12
Aim
Based on a comprehensive review of the literature above, statins appear to have a strong
mechanistic rationale supporting clinical findings of its anti-thrombotic effects in
preventing VTE, however when it comes to resolution of an incident VTE, only animal
trials have been conducted. This study hypothesizes that statins are associated with
greater thrombus resolution or improvement in humans. The aim of this study is to
investigate thrombus resolution and improvement in a group of patients treated for DVT
while on statin compared to patients not receiving statin. Thrombus resolution and
clinically significant recurrence of VTE will be correlated with statin use.

13
Methods:
Study design
This is a retrospective study examining consecutive adult patients (age ≥ 18) presenting
with newly diagnosed acute lower extremity DVT at an academic tertiary care center
from January 2013 through April 2014. The study protocol and waiver of informed
consent was approved by the human investigation committee. Patients were divided into
2 groups based on their use of statin therapy or lack thereof, and all patient records and
ultrasound reports were reviewed. Patients initiating statins after developing DVT were
considered to not have a standing history of statin use and were included in the non-statin
group for initial presentation but were analyzed with the statin group for long-term
outcomes. Patients who had continuous statin use prior to DVT incidence but
discontinued after were included in the statin group for initial presentation but were
analyzed with the non-statin group for outcomes.

Variables
Patient demographics and comorbidities as well as ultrasound characteristics were
extracted from the electronic medical records. Patients’ age, sex, history of
hypercoagulable disorders, hypertension, congestive heart failure (CHF), diabetes,
hyperlipidemia, coronary artery disease (CAD), peripheral arterial disease (PAD),
cerebral vascular disease (CVD), cancer (including metastatic disease where present),
prior DVT, prior pulmonary embolism (PE), and smoking were recorded. In addition to
the aforementioned variables, other risk factors for VTE required to calculate the Caprini
score were also extracted.33 These included concurrent presentation of varicose veins or
14
swollen legs, concurrent central venous access, history of inflammatory bowel disease,
sepsis within the past month, pneumonia, chronic obstructive pulmonary disease, long
distance travel, immobilizing plastic cast, multiple trauma, paralysis, major surgery,
elective major lower extremity arthroplasty, and hip, pelvis, or leg fracture. The
prescription of antiplatelets (aspirin, clopidogrel, prasugrel, dipyridamole, ticagrelor,
ticlopidine) and statins at the time of diagnosis was noted.

The DVT characteristics were noted as occlusive vs non-occlusive, proximal vs distal,
and provoked vs unprovoked. Proximal DVT included thrombi in the popliteal vein or
above, while distal DVT included thrombi found exclusively in the tibial veins. Cases
that had thrombi in both the popliteal vein or above and in the tibial veins were
considered proximal. Provoked DVT was defined according to the methodology of
Brownson et al. and included DVTs associated with trauma, recent surgery, sedentary
travel >4 hours, or confinement to bed >72 hours within 30 days of the event, or initiation
of oral contraceptives, or a central venous femoral line (if ipsilateral DVT).34 Associated
PE confirmed by imaging was classified by severity. High-risk PE was defined as >15
min of hypotension (<90 mmHg); intermediate-risk PE was defined by blood pressure >90 mmHg with accompanying right ventricular dysfunction (dilation or elevated N-
terminal pro brain natriuretic peptide >500 pg./mL or brain natriuretic peptide >90
pg/mL) or myocardial necrosis (troponin I >0.4 ng/mL or troponin T >0.1 ng/mL), and
finally low-risk PE was defined by the lack of these characteristics.35 Cases without
definitive imaging but where PE was highly suspected by the treating physician were also
noted.
15

The treatment of DVT was reviewed for the type of anticoagulation therapy [warfarin,
low molecular weight heparin (LMWH), or direct oral anticoagulant (DOAC)] and its
duration. Only full-dose anticoagulation therapies with INR goal of 2-3 or more were
noted. The placement of inferior vena cava (IVC) filters and administration of systemic
or catheter-directed thrombolytics were also noted.

Outcomes and Statistical Analysis
The long-term outcomes of mortality, DVT recurrence, PE recurrence, and VTE
recurrence were compared between the 2 groups in follow-up through summer of 2017.
The patients who underwent repeat lower extremity venous ultrasound after one month of
diagnosis were noted. The repeat ultrasound report was reviewed, and the radiologist
assessment of thrombus burden compared to the initial ultrasound was noted as:
progressed (worsened), stable, improved, or completely resolved. The number of repeat
ultrasound exams varied by patient and thrombus burden may have changed or resolved
prior to the final repeat ultrasound, so we noted the date and results of the earliest repeat
ultrasound after one month of diagnosis where no further change in thrombus status was
observed in subsequent exams. Patients who died during follow-up were excluded from
analysis of VTE recurrence. A multivariable analysis was then used to determine the
independent factors associated with thrombus improvement or resolution, VTE
recurrence, and mortality. Factors that were included in the model consisted of age, sex,
provoked vs unprovoked DVT, distal vs proximal DVT, history of DVT, history of PE,
16
smoking, anticoagulation therapy (warfarin, LMWH, or DOAC), antiplatelet therapy,
hypertension, CHF, diabetes, hyperlipidemia, CAD, PAD, CVD, and cancer.

Statistical analysis
Characteristics of the sample were summarized using descriptive statistics and expressed
as percentages (n) in the case of categorical variables and mean and standard deviation
for continuous variables. Bivariate analyses were conducted using independent samples t-
tests to compare means and chi-square tests to compare proportions. Multivariable
logistic regression was performed to compare study groups on a dependent bivariate
outcome while adjusting for sample characteristics. Statistical analyses were performed
using SAS version 9.4.

Attribution of work performed
Identification of consecutive adult patients presenting with DVT in the study period was
performed by Nancy Huynh MD, Anand Brahmandam MD, and Kirstyn Brownson MD,
including basic demographic information and characteristics of DVT such as location,
associated PE, anti-thrombotic treatment, and history of underlying hypercoagulable risk
factors (e.g. cancer, oral contraceptives, immobilization, long travel, surgery, et cetera).
This data was double-checked by the author Charles Hsu, who then additionally collected
information including comorbidities and past medical history, medication use (including
statin and antiplatelet use), timing of statin initiation relative to presenting DVT,
radiographic reports upon presentation and on subsequent follow-up, and long-term
follow-up of outcomes through summer of 2017 (mortality, VTE recurrence). The author
17
performed a comparison of radiographic reports between initial and follow-up encounters
to determine change in the status of thrombus burden of DVT. All statistical analysis was
done by Jesse Reynolds in collaboration with Cassius Iyad Ochoa Chaar MD MS and the
author. The author drafted the manuscript for submission to the Journal of Vascular
Surgery: Venous and Lymphatic Disorders and the manuscript was reviewed by Cassius
Iyad Ochoa Chaar MD MS, Alfred Lee MD PhD, Wassim Fares MD MSc, and Jesse
Reynolds MS.

18
Results:
Study population
A total of 818 patients with DVT were identified during the study period; 279 (34%)
were on statin therapy at time of presentation and 539 (66%) were not. There were 5
patients who were initiated on statin therapy and another 5 who discontinued statin
therapy shortly after the diagnosis of DVT; these were accounted for in the analysis. The
three most common statin agents were atorvastatin (10 to 80 mg daily), simvastatin (20 to
80 mg daily), and pravastatin (10 to 80 mg daily), which collectively accounted for ~90%
of all statins used. Patients in the statin group were significantly older (72 ± 13 vs 63 ± 17
years) and were more likely to have risk factors and manifestations of atherosclerosis,
including hypertension, CHF, diabetes, hyperlipidemia, CAD, PAD, and CVD. There
was no significant difference in sex, history of cancer, or history of smoking between the
2 groups (Table 1).

More patients in the statin group were on antiplatelet therapy (55% vs 20%). In the statin
group, 130 patients were on single agent antiplatelet therapy (122 aspirin, 8 clopidogrel)
and 24 were on dual agent antiplatelet therapy (21 aspirin and clopidogrel, 3 aspirin and
prasugrel or dipyridamole). In the non-statin group, 97 patients were on single agent
antiplatelet therapy (88 aspirin, 9 clopidogrel), and 10 were on both aspirin and
clopidogrel. One or more hypercoagulable disorders were found in 44 patients across
both groups, the most common including Factor V Leiden (n = 12), antiphospholipid
syndrome (n = 9), antithrombin III deficiency (n = 5), and Protein C deficiency (n = 5).
Patients in the non-statin group had a higher proportion of hypercoagulable disorder or a
19
history of prior DVT while the patients in the statin group had significantly higher mean
Caprini scores. Additionally, there was no significant difference in the history of prior PE
(Table 1).

DVT characteristics and treatment
There was a higher proportion of high-risk PE (24% vs 11%, P = .046) in the statin
group. The 2 groups showed no significant difference in provoked DVT, proportion of
proximal DVTs, or association with PE. The groups also showed no significant difference
in the type of anticoagulation received, duration of anticoagulation therapy, IVC filter
placement, and proportion treated by thrombolysis (Table 2).

Long-term outcomes
There was no difference in the overall duration of follow-up between the 2 groups (741 ±
569 days vs 710 ± 565 days, P = .46). The frequency of repeat ultrasound after one month
(39% vs 45%, P = .13) and the intervals at which ultrasounds were performed following
initial presentation were not significantly different (496 ± 408 days vs 530 ± 409 days, P
= .46). There was no significant difference in mortality (41% vs 37%, P = .32) or
resolution or improvement of thrombus on repeat ultrasound between the groups (85% vs
77%, P = .12). There was no statistically significant difference in recurrence of DVT
(15% vs 17%, P = .52), PE (2% vs 4%, P = .41), or VTE (16% vs 19%, P = .43) (Table
3). A subgroup analysis of the characteristics of patients who underwent repeat
ultrasound after one month was performed (Supplement Tables 1 and 2). In this
subgroup, there was no significant difference between the statin and non-statin patients in
20
recurrence of DVT, PE, VTE or mortality (Supplement Table 3). On the other hand, a
comparison of the patients who did not receive repeat ultrasound after one month to the
patients that underwent repeat ultrasonography showed significant differences and
selection bias. Patients who did not receive a repeat ultrasound were older, suffered from
more comorbidities, and were less likely to receive anticoagulation or catheter-directed
thrombolysis as treatment for DVT. They had significantly higher mortality, shorter
follow up, and less likelihood of having recurrent VTE (Supplement Tables 4, 5, and 6).
Furthermore, there was no significant difference in time to repeat US between the statin
and non-statin groups when stratified into groups based on thrombus resolution status and
time to repeat US < 1 year, 1 – 2 years, 2 – 3 years, 3 – 4 years, and > 4 years
(Supplement Table 7).

Multivariable analysis
On multivariable analysis, which adjusted for potential confounders including those
variables in Tables 1 and 2, statin therapy (OR 3.23, 95% CI [1.32 – 7.87]) and
antiplatelet therapy (OR 2.25 [1.05 – 4.81]) were found to be independently associated
with thrombus resolution or improvement on ultrasound. Older age (OR 0.97 [0.95 –
0.99]) was found to be associated with lower thrombus resolution or improvement. No
significant association was found between thrombus resolution or improvement and the
following factors: sex, provoked vs unprovoked DVT, distal vs proximal DVT, CAD,
history of cancer, or anticoagulation by warfarin, LMWH, or DOAC as compared with no
anticoagulation (Table 4).

21
Distal DVT compared to proximal DVT (OR 0.53 [0.35 – 0.8]), anticoagulation with
warfarin (OR 0.19 [0.12 – 0.3]) or DOACs (OR 0.16 [0.09 – 0.3]) as compared to no
anticoagulation, and antiplatelet therapy (OR 0.57 [0.37 – 0.87]) were all independently
associated with lower mortality. Age (OR 1.04 [1.03 – 1.06]), CAD (OR 3.12 [1.80 –
5.41]), and history of cancer (OR 4.26 [2.91 – 6.25]) were found to be independently
associated with mortality. Mortality outcomes were not associated with statin therapy,
sex, provoked vs unprovoked DVT, or anticoagulation by LMWH as compared to no
anticoagulation (Table 4).

Finally, there was no significant association of any variable with recurrent VTE (Table
4).

22
Discussion:
This study found that statin therapy is associated with thrombus resolution or
improvement after DVT, consistent with the literature on statin’s pleiotropic
cardiovascular effects beyond its known lipid-lowering action. However, the
retrospective nature of this study places significant limitations on our ability to draw
inference on causation. On univariate analysis, the statin group was found to have no
significant difference in thrombus characteristics on repeat ultrasound from the non-statin
group, but patients on statins had significantly more cardiovascular comorbidities and
were older. The lack of an overt difference in thrombus resolution or improvement
between the two groups could be attributed to the counterbalancing influence of older age
and possibly greater overall comorbidity impairing or otherwise adversely affecting the
ability of thrombus remodeling. Even though this effect of statins has been described in
animal models, the clinical correlation in humans has not been reported, to our
knowledge.5,6 As established in the introduction, statins have both significant anti-
thrombotic effects and anti-inflammatory properties on the vascular endothelium, which
may explain the enhanced thrombus resolution seen in this study.36 A similar impact was
reported in the arterial circulation in a case report describing a patient having an aortic
mobile atheroma that was treated with high dose statin and dual anti-platelets resulting in
complete resolution of the atheroma on serial echocardiography.32

This study found that the impact on thrombus resolution did not translate into a clinical
effect of decreased recurrent VTE. While this is consistent with the findings of Brækkan
et al., who found a modest but insignificant decrease in recurrent DVTs in patients on
23
statins, other studies showed a clearly protective pattern.4,37-39 Smith et al. found a
statistically significant lower risk (26%) of recurrent VTE with statin use. Key
differences in the cohorts may account for the differing results. Smith et al. studied a
larger patient population (n = 4,350) and had a longer mean follow up of 3.4 years. Even
though the incidence of recurrent VTE observed in their study (16%) was similar to that
of our study, the larger sample and longer duration of follow up likely allowed Smith et
al. to demonstrate clinical differences in VTE recurrence that our study did not have
enough power to resolve. Additionally, our current study had a high proportion of cancer
patients (40% vs 23% in Smith et al.) and a high mortality on follow up (40%) decreasing
further the likelihood of the cohort to demonstrate significant differences in VTE
recurrence.39 While the JUPITER randomized trial demonstrated the effect of
rosuvastatin on prevention of VTE in 17,802 patients followed for 1.9 years, the trial
excluded patients with cancer and cardiovascular disease and focused on a population of
“apparently healthy adults” with no history of VTE.2 In a retrospective study of 170,459
patients with cancer followed for up to 1 year, El-Refai et al. found that statins had no
statistically significant protective effects against cancer-associated VTEs in most cancer
types with the exception of DVT in leukemia and PE in colorectal cancer.40 Therefore it
is plausible that the high proportion of cancer in the current study could explain the lack
of association between statins and VTE recurrence.

Since the submission of the original manuscript for publication, additional studies have
been published in the literature exploring topics covered in this thesis. In a 2019 Japanese
study (n = 3027) from the COMMAND VTE Registry, Yoshikawa et al. found that