Choosing between Enoxaparin and Fondaparinux for the management of patients with acute coronary syndrome: A systematic review and meta-analysis

Background Enoxaparin and Fondaparinux are potential anticoagulants which are used peri-operatively in the management of patients with Acute Coronary Syndrome (ACS). We aimed to compare the adverse clinical outcomes which are associated with the use of these anticoagulants in patients who were treated for ACS. Methods Online databases (PubMed/Medline, EMBASE, Cochrane library) were searched for studies which compared differences in clinical outcomes observed with the use of enoxaparin and fondaparinux in patients who were treated peri-operatively for ACS. Statistical analysis was carried out by Revman 5.3 software with odds ratio (OR) and 95% confidence intervals (CI) as the analytical parameters. Results Seven studies with a total number of 9618 patients (mainly composed of non-ST elevated myocardial infarction/NSTEMI) were included. This analysis showed mortality to be similarly observed between enoxaparin and fondaparinux with OR: 1.05, 95% CI: 0.67–1.63; P = 0.84. Myocardial infarction (MI) and stroke were also not significantly different throughout different follow up periods. However, minor, major and total bleeding were significantly lower with fondaparinux (OR: 0.40, 95% CI: 0.27–0.58; P = 0.00001), (OR: 0.46, 95% CI: 0.32–0.66; P = 0.0001) and (OR: 0.47, 95% CI: 0.37–0.60; P = 0.00001) respectively during the 10-day follow up period. Even during a follow up period of 30 days or a midterm follow up, major and minor bleeding still significantly favored fondaparinux in comparison to enoxaparin. Conclusion In patients who were treated for ACS, fondaparinux might be a better choice when compared to enoxaparin in terms of short to midterm bleeding events. This result was mainly applicable to patients with NSTEMI. However, due to a limited number of patients analyzed, further larger randomized trials should be able to confirm this hypothesis.


Background
Enoxaparin and Fondaparinux are potential anticoagulants which are used peri-operatively in the management of patients with Acute Coronary Syndrome (ACS) [1]. The Fifth Organization to Assess Strategies in Ischemic Syndromes (OASIS 5) trial showed fondaparinux to reduce the rate of major bleeding and net clinical benefit including death, Myocardial Infarction (MI), stroke and major bleeding in comparison to enoxaparin [2]. However, results from the French Registry of ST segment elevation and non-ST segment elevation MI (FAST-MI) 2010 showed a similar rate of bleeding and mortality between fondaparinux and enoxaparin [3]. Even though, FAST-MI cannot be compared to the OASIS 5 trial which consisted of a very large number of patients, we would still like to confirm the results through a meta-analysis by comparing the adverse clinical outcomes which were observed with enoxaparin and fondaparinux in patients who were treated for ACS.

Data sources and search strategy
Online databases (PubMed/Medline, EMBASE, Cochrane library) were searched for studies (English publications) which compared differences in clinical outcomes observed with the use of enoxaparin and fondaparinux in patients who were treated peri-operatively for ACS by using the searched terms 'enoxaparin and fondaparinux and acute coronary syndrome'. Later in this search process, the terms 'percutaneous coronary intervention, heparin' and the abbreviations ' ACS and PCI' were also used. Reference lists of suitable articles were also reviewed for relevant publications.

Inclusion criteria
Studies were included if they satisfied certain major criteria which were: (a) They were randomized controlled trials or observational studies comparing enoxaparin with fondaparinux in patients who were treated for ACS or in patients who were undergoing percutaneous coronary intervention (PCI). (b)They reported adverse outcomes (cardiovascular and bleeding outcomes) as their clinical endpoints. (c)They involved relevant data which could be used in this current analysis.

Exclusion criteria
Studies were excluded if they did not satisfy certain major criteria which were: (a) They were systematic reviews, meta analyses, case studies or letter to editors. (b)They did not include patients with ACS. (c)They did not report the previously mentioned clinical outcomes. (d)They were duplicates of the same study or they were associated with the same trial or cohort.

Types of participants
This analysis mainly included patients with non-ST segment elevated myocardial infarction (NSTEMI), patients with unstable angina (UA) and a small percentage of patients with ST segment elevated myocardial infarction (STEMI).

Definitions, outcomes and follow ups
Endpoints which were assessed included: (e) Major bleeding: any type of major bleeding; (f ) All bleeding (major and minor bleeding combined) including thrombolysis in myocardial infarction (TIMI) defined major and minor bleeding [4].
The follow up periods were classified as: -Less than 10 days (including in-hospital follow up period); -30 days follow up period; -Mid-term follow up (6 months to 1 year) period.
These outcomes and follow up periods have been summarized in Table 1.

Data extraction and review
Information and data including the name of authors, year of article publication, period of patients' enrollment, number of patients collected from each group (enoxaparin and fondaparinux), type of study (RCT or observational study), baseline characteristics of the patients, the outcomes reported, the follow up periods, the medications used by the patients, and the number of events reported with enoxaparin and fondaparinux respectively, were independently collected/extracted by two authors (PKB and MS). Disagreements were resolved by discussion with the third author (JY). Since this is a meta- analysis, the PRISMA guideline [5] was followed and the bias risk among the trials was assessed with reference to the features which have been stated in the Cochrane Collaboration [6]. All the trials were rated as having a 'low to moderate' risk of bias.

Statistical analysis
The statistical analysis was carried out by Revman 5.3 software with odds ratio (OR) and 95% confidence intervals (CI) as the analytical parameters. Two simple statistical methods were used to assess heterogeneity [7] across the studies namely the Q statistic test and the I 2 test.
A P value of less or equal to 0.05 was considered as statistically significant whereas a P value greater than 0.05 was insignificant in this analysis.
On the other hand, a fixed effects model or a random effects model was used during the analysis depending on the value of I 2 . A fixed effects model was recommended if an I 2 value <50% was obtained, whereas if the I 2 value was >50%, a random effects model was recommended.
In addition, this I 2 value was also used to predict heterogeneity. The lower the I 2 value, the lesser would be the heterogeneity whereas heterogeneity would increase with an increasing I 2 value.
Sensitivity analyses were also carried out by excluding each study one by one and a new analysis was carried out each time.
Publication bias was visually observed by analyzing the funnel plots which were generated through the RevMan software. Since this analysis did not involve a large volume of studies (only 7 studies were available), asymmetry of the funnel plots was sufficient to represent publication bias. Normally, if the total number of studies which were included was more than 10, other methods would have been more appropriate.
Ethical approval was not required for this type of study.

Search results
Seven hundred and eighty-seven (787) articles were obtained using the above mentioned searched terms from electronic databases. After a careful check of the abstracts and titles by the same two authors (PKB and MS), 764 articles were eliminated (not related to this current idea). Twenty-three (23) full-text articles were assessed for eligibility. Among these 23 articles, further studies were eliminated because they were: -Systematic reviews, case studies or letter to editors (2); -Articles with data which were irrelevant for this current meta-analysis (1); -Articles which did not report adverse clinical outcomes as their endpoints (4); -Duplicates or studies which were associated with the same trials or cohorts (9).
Finally, 7 studies were selected and included in this analysis (Fig. 1).
General features of the studies which were included Five (5) observational studies and 2 randomized trials were included with a total number of 9618 patients (6587 patients were treated with enoxaparin and 3031 patients were treated with fondaparinux). Regions of patients' enrollment included France, China, Brazil, and Canada with an enrollment period ranging between years 2003 to 2015. The general features of the studies which were included have been listed in Tables 2 and 3.

Baseline features of the studies which were included
The baseline features (Table 4) were as follow: mean age 57.3-67.0 years, majority of the patients were male patients, other co-morbidities and smoking history were also reported in both groups. Overall, there were no significant differences in the baseline features between patients who were treated with enoxaparin and fondaparinux.
The OASIS 5 trial had a large number of patients when compared to the other studies. However, in order for the current results not to be influenced by the results obtained in the OASIS 5 trial, the proportion of patients obtained from the OASIS 5 trial was reduced in order to adjust to this current analysis. To be more clear, only the percentage of female patients (20-30%) from the trial were included when analyzing the 30-day outcomes and only 11 to 12% of patients were included when assessing bleeding events. It should be noted that the total number of patients which were extracted was not reduced. For example, '10 events of death out of 100 patients' was represented as '1 event of death out of 10 patients'. Therefore, it is basically the same thing. Table 5 summarized other anti-platelet and anticoagulant medications which were used by the patients. It can clearly be seen that almost all the patients were also being treated by aspirin and clopidogrel. Glycoprotein IIb/IIIa, prasugrel and unfractionated heparin were also being used by certain patients.

Sensitivity analysis
Sensitivity analyses were carried out. During the 30-days follow up period, excluding study Zhang2016 resulted in a non-significant result associated with major bleeding with OR: 0.53, 95% CI: 0.27-1.04; P = 0.06. During the midterm follow up period, excluding trial OASIS 5 resulted in a non-significant major bleeding with OR: 0.43, 95% CI: 0.16-1.14; P = 0.09. For the other results, no significant change was observed.
In addition, when patients with STEMI were excluded, and an analysis was conducted based on patients with NSTEMI, major and minor bleeding were still significantly higher with enoxaparin, with OR: 0.46, 95% CI: 0.32-0.66; P = 0.0001 and OR: 0.40, 95% CI: 0.27-0.58; P = 0.00001 respectively during a 10-day follow up period. During a 30-day follow up period, only minor bleeding was significantly higher with enoxaparin, with OR: 0.48, 95% CI: 0.27-0.85; P = 0.01. In addition, during the mid-term follow up, major and minor bleeding were both significantly higher with enoxaparin, with OR: 0.55, 95% CI: 0.30-0.99; P = 0.05 and OR: 0.51, 95% CI:

Publication bias
This current analysis included of a total number of 7 studies (2 randomized and 5 observational studies) that assessed all the clinical endpoints with minimal evidence of publication bias which was visually observed through the funnel plots obtained (Figs. 9, 10, 11). These funnel plots represented minimal publication bias. Possible reasons might be the presence of

Discussion
This analysis aimed to compare enoxaparin with fondaparinux. Current results showed both anticoagulants to have similar mortality and stroke rates. However, fondaparinux was associated with a significantly lower major and minor bleeding when compared to enoxaparin in majority of the bleeding subgroups which were analyzed.
Surprisingly, even if the use of glycoprotein IIb/IIIa was more in patients who used fondaparinux, a low bleeding was observed with this anticoagulant when compared to enoxaparin.
At this stage, it would be interesting to know the mechanisms of action of fondaparinux and enoxaparin. Fondaparinux is a factor Xa inhibitor and does not inhibit thrombin (IIa) [8]. Enoxaparin on the other hand, binds to antithrombin to form a complex molecule that can irreversibly inactivate clotting factor Xa and it has less activity against thrombin [9]. This is how these two anticoagulants work.
The Brazilian Registry Data which was a multicentered retrospective observational study including 2282 patients also supported the result of this current analysis which is similar to recently published data in international literature showing the superiority of fondaparinux to enoxaparin in the Brazilian population in terms of bleeding outcomes following PCI (40.1% of the patients in the fondaparinux group and 35.1% of patients in the enoxaparin group underwent PCI) [10]. Similarly, the OASIS 5 trial which was a double blinded randomized trial comparing fondaparinux with enoxaparin in 6238 patients who underwent PCI showed reduced bleeding events to be associated with fondaparinux without any increase in mortality [2].
On the other hand, the OASIS 6 trial (12,092 patients obtained from 447 hospitals in 41 countries around the globe) which compared fondaparinux with placebo or unfractionated heparin in patients with STEMI showed that in patients who were not undergoing PCI, the former was associated with a lower mortality and reinfarction without increasing stroke or bleeding events [11]. However, our analysis showed no significant difference in mortality or stroke, but with significantly lower bleeding events following invasive procedures when fondaparinux was compared to enoxaparin. Could it be the invasive procedure which contributed to these different results obtained from OASIS 5 and 6?
To provide an answer to this question, a small Chinese study with NSTEMI patients who underwent PCI, showed no statistically significant difference in bleeding outcomes between fondaparinux and enoxaparin. However, it should not be ignored that both groups were also additionally treated with tirofiban [12] and the total  Moreover, the French cohort of NSTEMI patients who were predominantly managed invasively did not show fondaparinux to be superior to enoxaparin in terms of bleeding outcomes [3]. However, similar to the results of this analysis, mortality was not significantly different between these two anticoagulants. Nevertheless, it was shown that during the propensity score matched cohorts, a lower number of patients used aspirin or clopidogrel before admission.
Finally, the fact that several factors such as comorbidities, age, use of anti-platelets and anticoagulants, as well as the dosage and whether these blood thinning agents were used before and after admission and the effect of plaque should all be taken into consideration when assessing bleeding risk and other adverse clinical outcomes in such patients. For example, a recent metaanalysis showed thin-cap fibro-atheroma to be highly associated with culprit plaque rupture [13]. This prevalence was higher in patients with STEMI compared to patients with NSTEMI or unstable angina. Therefore, clinical outcomes might vary from study to study due to the influence of such factors.

Novelty
This analysis is new in several ways: -It is among the first meta-analyses comparing enoxaparin with fondaparinux in patients who were being treated for ACS. -This interesting idea is very important clinically.
-A low to moderate level of heterogeneity was observed in several of the subgroups analyzing the outcomes, which might be a positive aspect of this analysis. -Randomized patients and patients obtained from observational studies were combined as well as separately analyzed.

Limitations
Limitations were as follow: -Due to the limited number of patients analyzed, the results might not be very accurate.

Conclusion
In patients who were treated for ACS, fondaparinux might be a better choice when compared to enoxaparin in terms of short to midterm bleeding events. This result was mainly applicable to patients with NSTEMI. However, due to a limited number of patients analyzed, further larger randomized trials should be able to confirm this hypothesis.