Aortic dissection is a life-threatening cardiovascular disease that causes ~10,000 deaths in the United States each year (21% before admission and 32% in-hospital) (1, 2). Recently, Sweden indicated that the incidence has increased about 7.2/100,000 (3). However, the early identification and diagnosis of high-risk chest pain as acute aortic dissection (AAD) is the major challenge. Treatment measures include coronary angiography and thrombolytic drugs, which may cause poor prognosis in 24.8% of AAD, who are misdiagnosed as acute myocardial infarction (AMI) or acute pulmonary embolism (PE) (4–6). Not only that, chest CT or MRI is time-money-consuming and limited by emergency room conditions compared to biomarkers for the diagnosis of AAD (7, 8).

Several researches have investigated AAD potential biomarkers for faster and more accurate clinical treatment, such as smooth muscle myosin (9), calcium binding protein (10), soluble elastin fragments (11), soluble ST2 (5), and D-dimer (12). While increased values for D-dimers raise suspicion for AAD (4), it is difficult to use it to desciminate AAD from PE in daily practice, and there are nine (2.43%) exhibited negative D-dimer results in 370 AAD patients according to our previous studies (13). Moreover, the younger and smaller thrombosis groups have low specificity. A valuable diagnostic marker should provide information for early identification or elimination to improve the clinical treatment of AAD (12).

Lysophosphatidic acid (LPA) is a small and simple glycerophospholipid (1-acyl-2-hydroxy-3-phosphoglycerol structure) (14), which is an early molecular marker of coagulation cascade activation and cardiovascular-injury (15, 16). Previous researches have shown that lysophosphatidylcholine (LPC) was also involved in aortic aneurysm formation and decreased substantially in AAD, which is hydrolyzed to produce LPA (17, 18). The LPA may be produced earlier than D-dimer, or have a better diagnostic performance than D-dimer and AAD risk scoring system (19, 20), but the relationship between LPA and AAD is unclear. The plasma LPA levels of patients with acute chest pain (AAD, AMI, PE) and healthy participants were measured and compared to evaluate the diagnostic performance in distinguishing AAD from other chest pains.

A total of 204 patients, including 86 AAD patients, 60 AMI patients, 28 PE patients, and 30 healthy participants (Normal), were selected (Figure 1). The patient characteristics after 48 h of onset are shown in Table 1 and Supplementary Table 2. AAD patients were mostly younger males with no significant difference in symptom onset time in all groups than non-AAD patients (Supplementary Table 3). The D-dimer and LPA levels were significantly higher in AAD patients than in non-AAD patients on the whole (Supplementary Table 2). LPA level (344.69 ± 59.99) was significantly different from AMI, PE, and Normal. However, AMI and PE were not statistically different (Figure 2A). D-dimer levels in AAD and PE were higher but no statistical difference between the two (Figure 2B). Logistic multiple regression analysis showed that the blood pressure differences, hypertension history, D-dimer, and LPA were independently associated with AAD (P < 0.001) (Table 2).

Pearson analysis showed that LPA level was positively associated with D-dimer, p < 0.05) (coefficient of 0.17 in AAD, 0.15 in AMI, and 0.24 in PE). There was no significant correlation in Normal (Supplementary Figure 2).

This study, which included 174 suspected AAD patients and 30 healthy participants, found that LPA may be used for the clinical diagnosis of AAD. LPA, mainly produced by activated platelets, may be an early biomarker of the initiation of thrombosis and coagulation. Researches have shown that coagulation cascade activates platelets to release a large amount of lysophospholipids and autotaxin with phospholipase D activity stored in alpha particles at the same time. The two substances undergo a biochemical reaction to produce LPA, which increases the plasma concentration (14, 25, 26). LPA binding to its receptors (LPA1-6) on platelet surface activates platelets and forms a positive feedback reaction, alters platelet in morphology, promotes aggregation and thrombus stability (27–29). Moreover, LPA activates LPA1 and LPA3 expressed in vascular endothelial cells (30), promotes immune response and aggravate endothelial cell damage via Gαi-RhoA-ROCK-NF-κB dependent pathways (31). LPA also increases the expression of intracellular matrix metalloproteinase-2, which will remodel the extracellular matrix and reduce the aggregation of endothelial cells to promote the migration and aggregation of inflammatory cells, thereby damaging endothelial cells to activate the coagulation cascade (32).

We found that the magnitude of elevated LPA can distinguish patients with AAD from patients with AMI, PE, and the healthy within 48 h after symptom onset. AAD is a disease with high mortality due to severe damage to the aortic structure. There are also some biomarkers showing clinical prognosis including CRP (33), NT-pro BNP (34) and cardiac troponin (35). When aortic dissection occurs, disruption to the aortic media immediately changes aorta hemodynamics, then intramural hematoma expand (especially when the intimal layer is also disrupted) as the blood flow inside media (5). The tissue coagulation factor III directly exposed to the blood in the arterial smooth muscle activates the exogenous coagulation pathway and promotes the coagulation cascade (13). Platelets, an indispensable role in the coagulation cascade, are continuously activated by the LPC-ATX-LPA pathway to produce more LPA before the aortic dissection is surgically repaired (14, 36). Although some studies speculate that the half-life of LPA is only 2–3 min under the action of LPPs, the continuous existence of disruption to the aortic media and positive feedback may be a reasonable explanation for the increase in plasma LPA in AAD patients (37, 38).

Our results suggested that the degree of the elevation of LPA levels is associated with the different magnitudes of vascular injury among AAD, AMI, and PE. Although LPA level in PE and AMI patients was higher than the healthy, there were still lower than AAD in our study. Substantial aortic hemodynamic changes significantly increased the circulating LPA level in the aorta (largest artery) compared with the small and medium blood vessels of pulmonary embolism and acute coronary syndrome. Studies have shown that LPA level is associated with the release location, releasing LPA directly and quickly to the aortic circulation may be another reason for the higher degree of AAD patients (39).

In addition, we speculated that LPA peaks at 12 h according to the symptoms of AAD patients, and our results also found that the threshold of LPA is 298.98 mg/dl, the specificity and AUC are 0.77 and 0.857, respectively, which are better than D-dimer (P < 0.05, Delong test). This means that as a marker of platelet activation, LPA ≥ 300 mg/dl indicates a high risk of AAD, and may be earlier than D-dimer. However, D-dimer has higher sensitivity and negative predictive value compared to LPA. As a fibrin degradation product in the circulation after thrombolytic fibrinolysis, D-dimer has been found to increase similarly in many diseases, including PE and AAD. In fact, D-dimer aids clinical diagnosis for PE only as a rule-out tool when the test result is negative (5).

To the best of our knowledge, this is the first study showing that LPA is associated with aortic dissection and could be a novel AAD biomarker. However, this study has some limitations. It is still unclear how LPA changes over time in AAD patients, a large-scale prospective multicenter study is needed to confirm the generalizability of the findings, the diagnostic validity, and the accuracy of this novel detection method since this was a single-center study. Second, the assay method used could provide an inconsistent absolute value for the LPA concentration, influencing the recommended cut-off level. Therefore, various detection methods should be accurately calibrated. Thirdly, this study did not include all undifferentiated chest pain patients, hence, patient selection could be biased. Finally, LPA <300 mg/dl is difficult to rule out AAD in daily practice, and other diseases (ovarian cancer) (40) can also increase LPA levels. Therefore, further confirmatory diagnosis based on medical images is essential to prevent misdiagnosis in clinical practice.

LPA showed superior overall diagnostic performance to D-dimer in early AAD diagnosis may be a potential biomarker, but additional studies are needed to determine the rapid and cost-effective diagnostic tests in the emergency department.

The data analyzed in this study is subject to the following licenses/restrictions: We are deeply sorry for this, because there is still some research in progress for the time being. In order not to affect the following research, it is not convenient to disclose the data. Requests to access these datasets should be directed to 188212324@csu.edu.cn.

The hospital institutional review board of the Second Xiangya Hospital approved the study. The data collection and analysis followed the Ethics Committee of the institution and the Declaration of Helsinki. The Ethics Committee of the institution reviewed the patient consents, and the data were used only for research purposes.

XP and XC: drafted, revised, and reviewed the article. YZ and GY: conducted statistical analysis. ZH, HZ, ZP, WP, and ND: reviewed and revised the manuscripts. TG and MZ: organized the database. All authors significantly contributed to the conception, study design, execution, data acquisition, analysis, interpretation, approved the final version, agreed on the journal, and are responsible for this study.

The Key Research and Development Program of Hunan Province; Hunan Health and Family Planning Commission Project; Hunan Provincial Natural Science Foundation Project (Nos. 2019SK2022; 20200063; 2021JJ30923) supported this study.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.