DISCUSSION
The results of this study demonstrate that there are predictable and quantifiable changes in ROTEM values during surgery in ATAAD and elective aortic surgery with CPB. Surgery significantly and negatively impacted all ROTEM values assessed in this study (EXTEM CT, INTEM CT, HEPTEM CT, EXTEM MCF, and FIBTEM MCF). The greatest impairment in coagulation parameters occurred consistently in patients with ATAAD. This study demonstrated that ATAAD caused an activation of the coagulation shown in ROTEM prior to surgery (T0), which developed to a coagulopathy during CPB (T1 and T2) and was not fully recovered compared to elective controls at wound closure (T3) despite significantly greater use of procoagulants and transfusions. Our ROTEM-guided transfusion protocol does not seem to catch the full need of transfusions as most of our patients received more transfusions than what ROTEM would suggest. However, bleeding volumes and the need for re-exploration for bleeding or tamponade did not differ and had favorable outcomes in both groups.
Aortic dissection leads to blood being exposed to tissue factor, extracellular collagen, and other subendothelial structures that activate the coagulation process. This is evident by the decreased MCF in FIBTEM indicating consumption of fibrinogen. MCF in EXTEM also was decreased but did not reach the level of significance. Both EXTEM and INTEM CT showed a trend towards longer clotting time indicating reduced amounts of clotting factors. Combined, this indicates an established activation of coagulation when ATAAD develops resulting in a consumption coagulopathy. During surgery, both groups showed similar trends in all ROTEM variables, but the ATAAD group had consistently more impaired values, which could indicate that the dissection consumed coagulation factors, platelets, and fibrinogen prompted by more profound hypothermia.
At the end of surgery, both groups showed similar ROTEM findings. The ATAAD group had longer HEPTEM CT than the control group, but not INTEM CT. This indicates a reduction of factors in the intrinsic pathway in the ATAAD group and a remaining heparin effect in the control group. In the ATAAD group, FIBTEM is normalized and equal to the control group. These findings are in line with similar work by Liu et al. and Guan et al. who did serial TEG analysis on patients with ATAAD21, 22. Data in both studies demonstrated that ATAAD initiates a consumption coagulopathy, and surgery affects fibrinogen and clotting factors more than platelets. However, TEG is not able to detect differences between intrinsic and extrinsic pathways. The studies lacked either a control group or did not provide differences between samples with heparinase.
Postoperatively, the patient was hypercoagulable in both groups, primarily in terms of FIBTEM, which may be explained by an increase in fibrinogen levels caused by inflammation 23. ROTEM at day 4-5 (T5) showed normal CT in INTEM and a prolonged CT in EXTEM in both groups. The MCF was elevated in FIBTEM and EXTEM. The increase is likely driven by high fibrinogen levels. The prolonged CT in EXTEM indicates a higher threshold for extrinsic pathway activation, likely induced by increased activity by inhibiting factors such as antithrombin, protein C, and protein S. This is supported by the normal PT-INR and increased antithrombin at day five as shown previously17.
Point-of-care testing with either ROTEM or tromboelastography (TEG) has been proven to reduce the need for red blood cell transfusion and reduce bleeding in cardiac surgery 10. The clotting time (CT) of INTEM and EXTEM test the same pathways as activated partial thromboplastin time (APTT) and prothrombin time (PT-INR), respectively. Maximum clot firmness (MCF) in both INTEM and EXTEM is an estimate of platelet number and fibrinogen levels while MCF in FIBTEM is primarily a measurement of fibrinogen levels 11. One of the main benefits of ROTEM are its fast results compared to compared to routine plasma-based laboratory coagulation tests (RLT) which enables serial testing of the coagulation and allows for faster response on changes in the coagulation during and after surgery, and our results generated by ROTEM are in line with previous reports using RLT 17, 24. However, one main difference between ROTEM and RLT is that RLT has well-established quality assurance programs with imprecision results, coefficient of variation (CV%) <5%, which is more difficult to achieve for a whole-blood system like ROTEM, reflected in the reference values that are wider in ROTEM counterparts. Another benefit of RLT is that ROTEM is harder to interpret and in the setting of complex coagulopathy, the use of both RLT and ROTEM provides a more nuanced picture 25.
To be able to use ROTEM as a substitute for RLT in a transfusion protocol it needs to adequately identify differences in coagulation seen in RLT. When comparing our ROTEM data with RLT in the same cohort, we find that ROTEM does not detect as many pathologies as RLT does. ROTEM identified all cases with low levels of fibrinogen, a finding supported by previous studies 13, 26. However, low levels of clotting factors and platelets seem to be underdiagnosed by ROTEM. In this study, ROTEM suggests that 35% of the patients require PCC or FFP, but RLT indicates that all patients have decreased levels of coagulation factors. This is also demonstrated by Rugeri et al. 26who showed poor correlation between CT and APTT/PT-INR. Platelet levels were also underdiagnosed by ROTEM, where only one out of seven thrombocytopenic patients were detected compared to RLT.
Our ROTEM-guided transfusion protocol was introduced in 2015. It follows a similar structure to previously published protocols9, 10, 12. The adherence to protocol in this study, however, was not always optimal. When analyzing the transfusions, all patients in the ATAAD group received platelets, fibrinogen, and PCC and/or FFP. Compared to the transfusion protocol, all patients in the ATAAD group met the criteria for PCC/FFP substitution, >80% for fibrinogen, but only 50% met the criteria for platelets and RBC at T2. This could be interpreted in different ways: either pre-emptive transfusions were used to prevent coagulopathy or the protocol was bypassed due to clinical coagulopathy not detected by the protocol. As mentioned earlier, ROTEM failed to identify the patients with low levels of coagulation factors or platelets indicating that the ROTEM-guided algorithm used in routine surgery may not be directly translated to ATAAD or other complex surgery. Also, our ROTEM does not contain platelet function analysis, which may explain why the algorithm was not followed for platelet transfusion.
Although ROTEM has been proven to reduce transfusions in cardiac surgery, some studies show no predictive value of ROTEM11, 14-16, 27. One potential reason might be its imprecision. In elective surgery, a preoperative test enables the patient to act as its own control, and changes in ROTEM could better be interpreted in that context. However, in the acute setting of surgery for ATAAD, the patients’ preoperative values are impaired, and preoperative levels have a limited value. This is supported by several studies of ROTEM and TEG showing that comparing pre- and post CPB levels is better at predicting bleeding rather than a post CPB ROTEM/TEG analyzed with predetermined cut-off values 11, 14-16, 27.
There are several limitations to this study. First, the sample size is not large leaving room for type II errors. However, including more subjects would require a longer inclusion period, and smaller differences found in a larger cohort may have questionable clinical value. The use of a control group undergoing aortic surgery is of benefit, but a better-matched surgical procedure with deep hypothermic circulatory arrest could possibly have given even more insights. Since adherence to the transfusion protocol was impaired, the true effect of a ROTEM-guided protocol could not be evaluated.