Statin use and 30-day mortality in patients with acute symptomatic pulmonary embolism

1 INTRODUCTION

The influence of statins on survival in patients with coronary, cerebrovascular, or peripheral artery disease has been largely reported.^1-5 In addition to lowering the low-density lipoprotein cholesterol levels, statins have pleiotropic effects, including anti-inflammatory, antithrombotic, and profibrinolytic properties. Statins reduce tissue factor expression and thrombin generation, and attenuate fibrinogen cleavage.^6-12

These factors make it plausible that statins may help in the treatment of patients with venous thromboembolism (VTE). In patients with VTE, some authors have found that statin use was associated with a decreased risk of VTE recurrences.^{13, 14} Moreover, in a population-based registry of patients with acute pulmonary embolism (PE), those using statins during follow-up (median, 199 days) had half the risk of dying than non-users.¹⁵ In two studies using the Registro Informatizado de Pacientes con Enfermedad TromboEmbólica (RIETE) registry, we also found that statin users at baseline had a significant decrease in all-cause mortality during follow-up (median, 177 and 192 days, respectively) compared to non-statin users.^{16, 17} Whether such an association is driven by short-term mortality in patients with acute symptomatic PE has not been thoroughly investigated, in part due to small sample size in prior cohorts.

We used the data from RIETE, a large multicenter, ongoing, international registry of patients with objectively confirmed, acute VTE (ClinicalTrials.gov identifier: NCT02832245). In the current analysis, we aimed to evaluate the association between statin therapy and short-term mortality, compared to no statin therapy. To account for biological plausibility of the potential findings, we decided for the current study to: (1) only include patients with acute symptomatic PE (we excluded patients with isolated deep vein thromboses [DVT]), (2) choose 30-day all-cause mortality as the primary outcome, (3) assess the strength of associations across the strata of intensities of statin therapy on mortality, (4) perform the analyses in multiple ways (multivariable, sensitivity, and propensity-score–matched analysis), and (5) use falsification endpoints to reduce the risk from findings being driven by residual confounding.

2 METHODS

3 RESULTS

From January 2009 to April 2021, 31 169 patients with acute symptomatic PE were recruited. Of these, 5520 (18%) were using statins: at low intensity: 829, intermediate intensity: 3636, at high intensity: 1055 (Figure 1). Compared to non-users, statin users were 10 years older and more likely to have hypertension, diabetes, chronic heart failure, prior myocardial infarction, prior stroke, peripheral artery disease, or CrCl levels <60 mL/min (Table 1). They also were more likely to be using antiplatelet drugs and had lower serum levels of total- and low density lipoprotein (LDL) cholesterol than non-users. Compared to patients using low-intensity statins, those on high-intensity statins were the same age but more likely to have chronic heart failure, diabetes, prior myocardial infarction, stroke, peripheral artery disease, or to be using antiplatelet drugs than those using low-intensity statins, and had the lowest levels of total- and LDL cholesterol in serum (Table 1).

The proportion of patients initially presenting with PE and hypotension (systolic blood pressure levels <90 mm Hg) was similar in statin users and non-users (Table 2). Statin users were slightly less likely to initially present with tachycardia, but more likely to have hypoxemia or raised troponin levels than non-users. The proportion of patients presenting with abnormalities on the echocardiogram (right ventricle dysfunction, raised pulmonary artery pressure levels or low tricuspid annular plane systolic excursion) was similar, but statin users scored worse on PESI (pulmonary embolism severity score) and simplified PESI than non-users. Most patients in all subgroups were initially treated with low molecular weight heparin (LMWH). Then, the proportion of patients receiving vitamin K antagonists, LMWH, or direct oral anticoagulants was similar in all subgroups (Table 2).

During the first 30 days, 198 patients (0.64%) developed VTE recurrences (recurrent PE: 123, DVT: 75), 544 (1.75%) suffered major bleeding (gastrointestinal: 139, haematoma: 149, intracranial: 75), and 1475 (4.73%) died (fatal PE: 255, fatal bleeding: 75; Table 3). The all-cause mortality rate was slightly lower in statin users than in non-users: 4.15% versus 4.86% (OR: 0.85; 95% CI: 0.73–0.98). The fatal PE rate was: 0.40% versus 0.87%, respectively (OR: 0.45; 95% CI: 0.29–0.69). The all-cause mortality rate according to the intensity of statin therapy was low intensity: 2.90%, moderate intensity: 4.26%, high intensity: 4.74%. Fatal PE rates were: 0.24%, 0.52%, and 1.04%, respectively. Compared to non-users, patients receiving low-intensity statins had a lower all-cause mortality rate (OR: 0.59; 95% CI: 0.39–0.88), those on moderate intensity had a non-significantly lower mortality rate (OR: 0.87; 95% CI: 0.74–1.03), and those on high intensity had a similar rate (OR: 0.98; 95% CI: 0.74–1.30) of all-cause mortality. The differences in all-cause mortality (OR: 0.32; 95% CI: 0.05–1.07) and in fatal PE rates (OR: 0.32; 95% CI: 0.02–1.62) between patients using low-intensity statins and non-users were highest during the first 48 hours (Figures 1 and 2).

On multivariable analysis, statin users had a significantly lower risk for all-cause death (OR: 0.65; 95% CI: 0.56–0.76) and for fatal PE (OR: 0.42; 95% CI: 0.28–0.62; Table 4). We repeated the multivariable analyses after replacing statin use (yes/no) for the three intensities of statin therapy. Compared to non-users, the risk for all-cause death was significantly lower in patients using low- (OR: 0.51; 95% CI: 0.34–0.77), moderate- (OR: 0.68; 95% CI: 0.57–0.81), or high-intensity statins (OR: 0.68; 95% CI: 0.51–0.92). As for fatal PE, only patients using low- (OR: 0.22; 95% CI: 0.06–0.90) or moderate-intensity (OR: 0.39; 95% CI: 0.24–0.63) statins were at a lower risk than non-users.

Adjustment used a mixed effects logistic regression model to take into account the possible random effect of the hospitals; the previous results did not have a substantive change (Table 4). Further, there was no association between the use of statins and cancer-related mortality (OR: 0.87; 95% CI: 0.65–1.22).

In the subgroup analyses, we did not find differences in the estimates of the effect of statins in subgroups according to systolic blood pressure levels, active cancer, or use of thrombolytic therapy. All interactions were statistically non-significant (Figures 3 and 4).

4 DISCUSSION

Our data, obtained from a large series of consecutive patients with acute symptomatic PE, confirm that statin users were 10 years older and more than 3-fold more likely to have arterial hypertension, diabetes, prior artery disease, or to be using antiplatelet drugs than non-users. The severity of the PE (according to PESI or simplified PESI) scored worse in statin users, and their treatments were similar. However, the 30-day rates of all-cause death (crude OR: 0.85; 95% CI: 0.73–0.98) or fatal PE (crude OR: 0.45; 95% CI: 0.29–0.69) were significantly lower in statin users than in non-users. Compared to non-users, the older age, the higher proportion of patients with comorbidities, the worse severity of the initial PE event, and the similar therapy were common in all three subgroups of patients using different statin intensities. Multivariable adjusted analyses confirmed that the lower risk for all-cause death or fatal PE persisted after adjusting for several potential confounders. The lower risk for all-cause death consistently appeared in patients using low- (51%), moderate- (68%), or high-(68%) intensity statins. The 33% reduction in the adjusted risk for 30-day mortality in our study was consistent with the 47% reduction found in a Dutch study in which PE patients were followed-up for a median of 199 days,¹⁹ and our two prior studies in VTE patients followed-up for more than 100 days.^{16, 17} Of note, although we noted a significant reduction in PE-related mortality, there was no significant change in the falsification endpoint, that is, cancer-related mortality.

This is the third study on the association between statin use and mortality in patients with VTE using the RIETE registry. The first study included 32 062 patients with a first episode of VTE (either DVT or PE) followed-up for a median of 177 days.¹⁶ During the course of anticoagulant therapy, statin users and non-users had a similar risk for VTE recurrences or major bleeding, but users had a lower risk for death (HR: 0.62; 95% CI: 0.48–0.79). The second study validated our prior findings (HR: 0.68; 95% CI: 0.59–0.79) in a subsequent cohort of 19 557 VTE patients followed-up for a median of 192 days.¹⁷ In the current study, we focused only on patients with acute PE, and limited follow-up to the first 30 days because in RIETE we do not gather information on how many patients using statins at baseline kept receiving statins after discharge. Again, statin users had a significantly lower risk to die compared to non-users. These data are fascinating but should be taken with caution. They must be confirmed in specifically designed randomized trials before considering statins an additional therapy for patients with acute PE.

The exact mechanism underlying any protective effects of statins on the risk for death in patients with acute PE is unknown. Statins inhibit the enzyme 3-hydroxy-3-methylglutaril coenzyme and decrease serum levels of cholesterol. In addition to these effects, statins have pleiotropic mechanisms such as platelet inhibition, reduction of inflammation and C-reactive protein, increased production of nitric oxide, or down-regulation of the coagulation cascade.^{7, 20-24} Other authors have suggested that the beneficial effect of statins beside the cholesterol-lowering effect may be due to the fact that persons who receive treatment with statins have a healthier lifestyle and/or a lower risk profile.^{25, 26}

Data from RIETE reflect routine, unmonitored medical practice involving a broad spectrum of patients with VTE. It can, therefore, provide insights into the natural history of VTE and to be hypothesis generating. However, it has several limitations that need to be addressed. First, this study is an observational study with no randomization to statin versus no statin users. Thus, some non-measured variables, such as the reason for receiving statins, or patient preference or socioeconomic factors impacting the use of statins have not been considered. In this study, we performed extensive adjustment for several variables that may have impacted the use of statins, and the results remained substantively similar. Nevertheless, residual confounding may remain, as certain potential confounding variables may have not been available or may have not had the desired level of granularity. Third, in RIETE we only gather information on the use of statins at baseline, and have no information on adherence afterward. However, the length of follow-up was only 30 days; differences in mortality started to appear during the first few days and anticoagulant therapy is not a contraindication to statin use. Our falsification endpoint analysis did not indicate an association between statin use and short-term cancer-related mortality. This observation, although not conclusive, makes the possibility of residual confounding as the only explanation of our findings less likely. Fourth, uncontrolled healthy-user or healthy-adherer bias cannot be excluded either. However, statin users were disadvantaged for mortality considering that they were, on average, 10 years older and were more likely to have severe comorbidities than non-users. Finally, we failed to find differences among patients receiving different intensities of statin therapy, maybe because patients receiving higher doses could be sicker at baseline.

In conclusion, our findings confirm that PE patients using statins at baseline are at a lower risk for all-cause death and for fatal PE within the first 30 days. This decrease was independent of different intensities of statins. Intervention studies specifically designed to confirm our findings are warranted.

INFORMED CONSENT

On behalf of all the authors of the publication “Statin use and 30-day mortality in patients with acute symptomatic pulmonary embolism,” I declare that all the authors have read, critically reviewed the paper, and agree to its publication.

CONFLICTS OF INTEREST

Dr. Carmine Siniscalchi received a speaker’s fee for congress presentation by MediK. Dr. Behnood Bikdeli reports that he is a consulting expert, on behalf of the plaintiff, for litigation related to two specific brand models of IVC filters. The other authors disclose no potential conflicts of interest.

AUTHOR CONTRIBUTIONS

Carmine Siniscalchi and Manuel Monreal conceived and designed the study. Carmine Siniscalchi, Alfonso Muriel, Behnood Bikdeli, José María Surinach, David Jimenez, Josè Luis Lobo, Cristina Amado, Aida Gil-Diaz, Egidio Imbalzano, and Manuel Monreal either performed and/or interpreted the tests and data. All authors revised the manuscript critically for important intellectual content and gave final approval of the manuscript submitted.