This study examined the usefulness of combination therapy with AT and rTM for septic DIC. The articles analyzed were all observational studies. Based on our study, combination therapy tended to improve mortality, although there was no statistical difference in mortality. There was also some concern that the combination of anticoagulant AT and rTM would increase bleeding complications. However, the results of this reviews suggest that bleeding complications do not increase with combined therapy. Heterogeneity amongst the included studies was also high.

This is the first systematic review and meta-analysis examining the efficacy and adverse events of AT + rTM for septic DIC. As no prior RCT has examined the effect of the combination therapy, the studies included were all observational studies.

In six studies that examined mortality, three presented results in terms of hazard ratios, and another three presented results in terms of adjusted odds ratios. In each of these studies, the combination of AT and rTM tended to reduce mortality compared with monotherapy, but the differences were not statistically significant.

Among the HRs examined, Sawano showed that combination therapy was particularly effective [20]. This single-center retrospective study included 111 patients (60 receiving monotherapy and 51 receiving combination therapy). One possible reason for the better results in combination therapy was the unevenness of patient distribution. The combination therapy group included more cases from 2009, whereas the monotherapy group (AT monotherapy) included more cases from 2006–2008, prior to the launch of rTM.

Iba et al. performed a similar study showing the effectiveness of combination therapy [13]. The study was a multicenter post-marketing study of AT consisting of 258 patients (129 monotherapy and 129 combination therapy).

These two studies showed significantly lower 28-day mortality in patients treated with combination therapy. Meanwhile, Umemura et al. [16] examined in-hospital mortality in a multicenter retrospective cohort study conducted in 42 ICUs in Japan with 808 patients and reported similar mortality in the combination therapy group and monotherapy group. However, both groups did not exhibit equal disease severity, and the combination therapy group was observed to have higher SOFA scores and DIC scores.

Among those studies examined with adjusted ORs, Iba’s study [14] analyzed 459 patients (monotherapy with AT 372 and combination therapy, 87) and found an improved prognosis with combination therapy. Suzuki [15] utilized the Diagnosis Procedure Combination database in Japan and constructed a matched pair of 378 patients with pneumonia-based septic DIC treated by anticoagulants (189 each, rTM monotherapy group and combination therapy group). In this study, although the difference was not statistically significant, the combination therapy group demonstrated lower mortality (40.2% vs. 45.5%).

Umegaki [25] utilized DPC data and examined the effect of combination therapy in 2222 patients (1017 in monotherapy with AT and 1205 in combination therapy). Again, the superiority of the combination therapy was not confirmed (OR: 0.97, 95% CI 0.78–1.21; P = 0.81). In this study, patients with septic DIC and with ventilator management were included, but the deviation of severe cases was not mentioned.

The present meta-analysis included highly heterogeneous studies (hazard ratio I2 = 60%; adjusted OR = 72%) with very different effect sizes across studies. Furthermore, we could not integrate all the studies due to the mixture of outcomes reported in HR and OR. However, both meta-analyses indicate that combination therapy tends to improve prognosis. Considering the mechanism of combination therapy, since antithrombin binds irreversibly to thrombin, it is suggested that AT administration may attenuate the APC-producing effect expected due to rTM by blocking the binding of thrombin to rTM. However, the clinical data in this study suggest that this view may not always be the case.

Since the studies included were all observational studies, there were some critical limitations. First, the treatment selection was unclear and was decided by the physicians in most of the studies. Although we were unable to confirm this, since severe cases were generally treated with combination therapy, it is unlikely that the combination therapy included more less-severe cases. Second, the treatment regimen was not consistent. The order of AT or rTM, whether given concomitantly or sequentially, and the time intervals between treatments were not clearly specified.

Owing to statistical issues, heterogeneous treatment regimens, and the lack of high quality, it is impossible to draw conclusions from the present study. However, we performed a systematic survey of the presently available data and observed that almost all the studies tended to show the beneficial effect of combination therapy. Therefore, we believe that combination therapy is potentially superior to monotherapy. Additionally, combination therapy has been indicated to be more effective in patients with severe thrombocytopenia and AT deficiency [26].

Wada et al. also reported that combination therapy may be useful for patients with low antithrombin and low fibrinogen [27]. High-quality observational studies and RCTs are necessary to make a recommendation in the future.

In patients with sepsis-related coagulation disorders, the consumption of coagulation factors and platelets results in a bleeding tendency [23, 24]. Therefore, bleeding complications due to anticoagulation therapy are the main concern of clinicians. In the previous studies, the incidence of bleeding was sufficiently low, with AT and rTM used individually [14]. However, the risk of bleeding may increase when both anticoagulants are combined. In this study, the increase in bleeding complications was not observed in combination therapy. However, the effect sizes of the two studies differed significantly, and since the studies were moderately high heterogeneous (I2 = 55%), the quality of the evidence was low.

The three studies used in the analysis are two large Japanese studies and a multicenter post-marketing survey [14, 16]; we therefore consider our results to be reliable. Although regarding bleeding complications integrated results suggested that combination therapy is not inferior to monotherapy, this result should be interpreted with caution due to the moderately high heterogeneity among the studies.

With regard to DIC withdrawal rates, three studies were eligible. However, since they were not adjusted by confounders, the variability in the results was large, and there was a large amount of missing data; therefore, we thought that it might not be appropriate to perform a meta-analysis. The large number of missing data may be a result of the impossibility of collecting the necessary data to assess DIC withdrawal since all the studies included in this analysis were observational studies. However, the rate of recovery from DIC is an important clinical item, as it is one of the key indicators to assess the effectiveness of treatment and prognosis. Therefore, high-quality observational and prospective studies should be carried out in the future to examine the effect of combination therapy on DIC withdrawal.

Japanese Clinical Practice Guidelines for Management of Sepsis and Septic Shock recommend the administration of AT or rTM for DIC. In clinical practice in Japan, AT and rTM combination therapy has been used at some facilities, and there have been various reports on the usefulness of this therapy [13,14,15,16]. However, there is no consensus on the usefulness of combination therapy, and there are no guidelines discussing combination therapy, which is why this study was conducted.

The balance of the apparent benefits and harms of combination therapy for the treatment of septic DIC patients found in this study suggests that there is validity for its clinical use.

Our study characterized study bias using the ROBINS-I assessment since all studies on combination therapy were nonrandomized studies, allowing for a more nuanced understanding of the study findings and similar publications in the field. Moreover, all the studies included in this review were conducted in Japan with very high heterogeneity. Thus, the generalizability of our findings to other countries remains uncertain.

Moreover, while there may have been no published studies with negative results or ineffectiveness regarding the combination therapy of AT and rTM, the possibility of the existence of published papers showing its effectiveness cannot be ruled out. In addition, although the effects and adverse events of combination therapy should be examined and compared to those of patients treated without anticoagulant therapy, we were unable to find such a study. Therefore, we compared the efficacy of combined therapy to that of either AT or rTM monotherapy.

Besides the seven studies discussed in this report, two other studies compared combination therapy with monotherapy [28, 29]. However, since the results were shown only with Kaplan–Meier curves, OR and HR were not presented, and these studies were excluded from the analysis. Furthermore, although these studies reported favorable outcomes on combination therapy, their results might differ if additional data were available.