Unfractionated Heparin Improves the Clinical Efficacy in Adult Sepsis Patients

A Systematic Review and Meta-analysis

Sifeng Fu; Sihan Yu; Liang Wang; Xiaochun Ma; Xu Li


BMC Anesthesiol. 2022;22(28) 

In This Article


In this study, we demonstrated that UFH was an effective treatment for sepsis. The 28 d mortality was relatively reduced by 16% in the UFH group. What's more, the 28 d mortality of sepsis patients with APACHE II > 15 was relatively reduced by 17% in the UFH group. There were no significant bleeding complications in UFH group, which indicated the safety of UFH. To our knowledge, our study is the first meta-analysis to focus on the effect of UFH in sepsis.

Almost all the sepsis patients experienced coagulation abnormalities, ranging from minor changes that could only be detected by extremely sensitive tests to DIC.[8] Many anticoagulants have been examined to ameliorate the mortality in recent years, such as the tissue factor pathway inhibitor (TFPI), APC and thrombomodulin (TM).[39–41] However, none of them has been proven to be effective in sepsis. Recently, a SCARLET RCT showed that administration of a recombinant human thrombomodulin (rhTM) did not significantly reduce 28 d all-cause mortality among patients with sepsis-induced coagulopathy compared with placebo.[42] The reasons for the negative results may be multifactorial, including long study period, large differences in the number of enrolled patients from different ICUs, concurrent using of heparin, long time interval from diagnosis to the first dose of rhTM and selected end-point. Therefore, this does not mean that anticoagulant therapy is ineffective. The results of this research ought to be interpreted critically.

The role of UFH in sepsis is much more than just an anticoagulant.[43] Our team have identified the beneficial effects of UFH clinically.[33,34] The crucial role of UFH in sepsis was further demonstrated in vivo and in vitro. UFH prevented lethality in endotoxemic mice.[44] UFH interfered with Krüppel-like factor 5 (KLF-5) mediated nuclear factor-κ (NF-kB) activation and contributed to the inhibitory effects of chemokines and monocytes migration.[45] Concurrently, UFH enhanced endothelial barrier function and angiopoietin (Ang)/Tie2 axis probably represented one of the mechanisms by which UFH exerted its protective effect.[46] A therapeutic dose of UFH could also protect glycocalyx from shedding by inhibiting inflammation.[47]

The conclusion of this meta-analysis was consistent with previous reports. Multiple lines of evidence suggested that UFH may improve clinical outcomes in sepsis. UFH is more suitable in China as well because it is widely available and inexpensive. Higher quality evidence is needed to guide clinical practice.[48]

To date, there were three meta-analyses on the role of heparin in sepsis published in English. Wang et al's meta-analysis published in 2014 concluded that heparin therapy reduced 28 d mortality in adult severe sepsis patients.[13] There were four main differences between the design of two studies. 1. Both the RCTs and NRCTs were taken into consideration in Wang et al's research. We only contained the RCTs. 2. Wang et al's research included trials on both UFH and LMWH as the intervention. We only analyzed the effects of UFH. 3. The patients who received continuous renal replacement therapy (CRRT) were excluded in our research since the use of anticoagulants during CRRT might affect the results. Wang et al's research did not exclude such patients. 4. We included studies from 2006 to 2021.

Subsequently, Zarychanski et al's meta-analysis was published in 2015, which showed that heparin was associated with decreased mortality in patients with sepsis, septic shock and DIC.[14] The distinction of two studies contained the following two aspects: first, Zarychanski et al's research involved trials relating to LMWH and other anticoagulants, but UFH was the only intervention in our research. Second, we drew the conclusion from literature within 15 years. Zarychanski et al' research involved literature from 1983 to 2014, so the heterogeneity was obvious.

The third meta-analysis was about the efficacy and safety of LMWH in patients with sepsis.[15] There are three differences from our research. First, in contrast with Yu et al's team, both English and non-English researches were included in our study which could decrease potential publication bias, while only Chinese studies were analyzed in Fan et al. 's research. Second, the larger sample the clinical trial contained, the more representative the research's conclusion was. The number of patients enrolled in our research was almost four times than that in Fan et al's research. Third, Fan et al's research evaluated the effects of LMWH in sepsis, whereas we studied the role of UFH. The action mechanisms of the two heparins are different, UFH seems to be more promising.

There are several advantages in this article. First, we enrolled studies both in English and in Chinese in this research, which meant the generalization of the current findings. Second, to our knowledge, our study is the first meta-analysis to focus on the effect of UFH in sepsis. Last, the large number of subjects included and the diversity in patients' characteristics provided diverse information.

The study also has several limitations. First, as shown in the forest plot, the participants in Levi et al's study took up more than half of the total number and all of them received recombinant activated protein C, which has been withdrawal from the market. Second, due to the various publication years, the definitions of sepsis, septic shock and DIC had changed. In 2016, the Sepsis-3 Task Force updated the clinical criteria for sepsis, excluding the need for systemic inflammatory response syndrome (SIRS) and the concept of severe sepsis.[7] Third, bleeding and other adverse events were incompletely reported which may influence the accuracy of results. For example, there were only four of fifteen trials reported bleeding information. Fourth, the doses of UFH and the treatment duration varied. Last, overall quality of the body of evidence was not high enough and there was a lack of multicenter RCTs.