Background

Oral cancer is one of the most common malignant tumours in the head and neck region, which accounts for approximately 3% of malignant tumour [1, 2]. More than 90% of oral cancers are oral squamous cell carcinoma (OSCC), which are prone to regional cervical lymph node metastasis [3, 4]. To date, the methods for the diagnosis and assessment of OSCC are still insufficient, so more in-depth exploration of biomarkers is needed.

The relationship between malignant tumours and coagulation was proposed as early as the 19th century by Trousseau [5], and has since been widely discussed. Lu et al. reported that pretreatment thrombocytosis is an independent predictor of poor prognosis in patients with OSCC and is closely associated with aggressive tumour growth, lymph node metastasis, and distant metastasis [6]. A previous study by our group revealed that an elevated preoperative D-dimer level was a poor prognostic factor for OSCC progression [7].

Thromboelastography (TEG) technology was invented by Hartert in 1948 [8]. This technology combines the physical properties of blood, such as viscoelasticity, by real-time detection of whole-blood samples and provides the dynamic changes in the whole coagulation-fibrinolysis process. Previous studies have shown significant correlations between TEG and conventional coagulation assays in the assessment of thrombotic and haemorrhagic risk in deep venous thrombosis (DVT) formation, chronic liver disease, fractures in elderly individuals, and postpartum haemorrhage [9,10,11]. However, analyses of TEG parameters among OSCC patients are lacking.

The aim of our study was to investigate the coagulation characteristics of OSCC patients, and their trends at various stages of tumour progression by analyzing the preoperative conventional coagulation indices and TEG parameters of OSCC patients.

Methods

Patient cohorts

We retrospectively collected data from OSCC patients and nonmalignant patients who visited the Department of Oral and Maxillofacial Surgery, Hospital of Stomatology, Sun Yat-sen University from January 2021 to July 2022. Informed consent was obtained. This study was approved by the Ethics Committee of Hospital of Stomatology, Sun Yat-sen University. All patients were aged between 20 and 80 year. All the OSCC patients were diagnosed by pathology. The control group included those who were hospitalized for other nonmalignant diseases. The exclusion criteria for patients were as follows: (1) history of coagulation system disorders or recent anticoagulant medication; (2) with other malignancies; (3) previous oncological treatment such as radiotherapy; or (4) menstruation, infection or inflammatory states, other severe haematologic diseases, or hepatic insufficiency.

Clinical data collection

Data on the demographic characteristics and clinical information of each patient were collected. Tumour TNM staging followed the 8th edition of the American Joint Committee on Cancer (AJCC)/International Union Against Cancer (UICC) Oral Cancer Staging Guidelines. Data on the preoperative coagulation indices of the above patients, including the platelet count (PLT), D-dimer level, activated partial thromboplastin time (APTT), prothrombin time (PT), thrombin time (TT), fibrinogen (FIB) and preoperative TEG parameters, were obtained. TEG data were detected and acquired via the TEG® 5000 Hemostasis Analyzer System (Haemonetics Corporation, USA). All the blood samples were performed by professional technicians from the department of clinical laboratory of the hospital according to standard operating procedures. The above coagulation test data were retrospectively collected from the medical records system. Thromboelastography is now widely used in clinical practice and, as a routine preoperative monitoring method in our hospital, it provides a more comprehensive assessment of the patient’s coagulation status, thereby reducing the risk of perioperative coagulation-related complications.

Meanings of TEG parameters

The parameters and clinical significance of TEG are summarized as follows: (1) R time (min) indicates the time from the start until the beginning of clot formation, representing the level and function of clotting factors; (2) K time (min) and Angle (°) respectively indicate the time (K) and speed (Angle) from the clot formation to when the clot reached an amplitude of 20 mm, representing the efficiency of the early stages of clot formation; (3) MA (mm) indicates the amplitude at the point of maximum clot strength, representing mainly the level and function of platelets; (4) LY30 (%) and EPL (%) indicate the rate (LY30) and the predicted rate (EPL) of clot dissolution after reaching the MA; (5) G (d/sc) indicates a clot strength score calculated on the basis of instrument-measured data; and (6) coagulation index (CI) indicates a comprehensive coagulation status.

Statistical analysis

With SPSS 25.0 software, each coagulation parameter of the primary OSCC group, recurrent/metastatic OSCC group, and control group was analyzed by means of the Kruskal-Wallis H test for comparisons among the three groups and by means of the Mann-Whitney U test for comparisons with the control group. The relationships between the preoperative coagulation parameters and the tumour stage of primary OSCC group was analyzed by means of the Mann-Whitney U test. The predictive value of coagulation indices for cervical lymph node metastasis was analyzed by means of the receiver operating characteristic (ROC) curve. P < 0.05 indicated statistical significance.

Results

OSCC patients presenting with preoperative hypercoagulability

A total of 311 patients with primary OSCC, 44 patients with recurrent/metastatic OSCC and 71 patients in the control group were included in the study (Fig. 1). No distant metastasis was found in our cohort.

Fig. 1
figure 1

Flowchart of patient cohort inclusion and exclusion

Full size image

Among the 71 nonmalignant patients in the control group, 30 had benign jaw lesions, 9 had localized oral inflammation, 8 had benign salivary gland disease, 8 had oral leucoplakia (including mild–moderate dysplasia), 7 had benign oral tumours, 4 had dentofacial deformities, 2 had vascular malformations, 2 had impacted teeth, and 1 had dry eye syndrome.

As shown in Table 1, the baseline sex, age and body mass index (BMI) did not significantly differ between the OSCC patients and the control group. However, among OSCC patients, the proportions of patients with a history of tobacco or areca nut use were higher.

Table 1 Demographic characteristics of the OSCC patients and the control group
Full size table

We compared the conventional coagulative indices and TEG parameters of the primary OSCC (primary group) and R/M OSCC (R/M group) to those of the control group (Table 2). The three-group comparison revealed significant differences in preoperative PLT, D-dimer and FIB values (P < 0.05). The TEG parameters, including K time, Angle, MA, G and CI, also differed among the groups (P < 0.05). Compared with that in the control group, the K time was lower, and the Angle was higher in the primary group. MA was increased in both the primary and R/M groups. The comprehensive coagulation scores, including G and CI scores, also increased. These findings indicate that patients with primary and R/M OSCC patients have a preoperative hypercoagulable state.

Table 2 Comparison of preoperative coagulation indices between groups
Full size table

Coagulation-fibrinolytic characteristics are correlated with tumour stage

To further explore the relationship between coagulation-fibrinolytic status and tumour stage, we compared coagulation indices across different tumour stages in primary OSCC patients.

We found that with increasing pathological T stage (pT stage), the preoperative PLT, D-dimer, and FIB increased, and TT decreased. With respect to the TEG parameters, the R time and K time decreased, and the Angle, MA, G, and CI increased, indicating that hypercoagulability became more obvious with increasing pT stage (shown in Fig. 2).

Fig. 2
figure 2

Correlations between preoperative coagulation parameters and the pT stage of primary OSCC patients. A-D, Conventional coagulation indexes. E-J, TEG parameters. T1 stage, n = 41. T2 stage, n = 117. T3 stage, n = 87. T4 stage, n = 66. All referring to pathological staging. The Kruskal-Wallis H test were used, P < 0.05*, P < 0.01**, P < 0.001***

Full size image

With increasing pathological N stage, only preoperative D-dimer (P < 0.001) and FIB (P = 0.002) levels were elevated (Fig. 3-A, B). In terms of the TEG parameters, the K time (P = 0.019) decreased, and the Angle (P = 0.031) was increased (Fig. 3-C, D). There were no significant differences in the other TEG parameters (P > 0.05).

Fig. 3
figure 3

Correlations between preoperative coagulation parameters and the pN stage of primary OSCC. A-B, Conventional coagulation indexes. C-D, TEG parameters. N0 stage, n = 196. N1 stage, n = 40. N2 stage, n = 40. N3 stage, n = 35. All referring to pathological staging. The Kruskal-Wallis H test was used, P < 0.05*, P < 0.01**, P < 0.001***

Full size image

Predictive role of coagulation indices in lymph node metastasis

We further evaluated whether coagulation parameters can predict cervical lymph node metastasis (pN stage). Among 311 patients with primary OSCC, 196 were pN0 stage and 115 were pN1-3 stage. The cut-off value of each variable was calculated with the Youden index (Supplementary Table 1). ROC curve analysis revealed that preoperative D-dimer (> 383ug/L), FIB (> 3.575 g/L), R time (< 5.95 min) and K time (< 1.75 min) were predictive of cervical lymph node metastasis, with the area under curve (AUC) of 0.660, 0.637, 0.569 and 0.571, respectively (Fig. 4).

Fig. 4
figure 4

ROC curves of preoperative coagulation parameters for the prediction of the pN stage of OSCC. A-B, Conventional coagulation indices. C-D, TEG parameters. Patients with primary OSCC (N = 311) were categorized as pN0 (n = 196) or pN1-3 (n = 115), and ROC curve analysis was used to predict lymph node metastases. AUC, area under curve. P < 0.05*, P < 0.01**, P < 0.001***

Full size image

Among the 241 clinical N0 (cN0) patients, 66 patients (27.39%) presented with pathologically confirmed cervical lymph node metastases, which led us to consider whether coagulation indices are predictive of occult metastases. As shown in Fig. 5, among the cN0 patients, preoperative D-dimer (> 392.5ug/L) and FIB (> 3.575 g/L) were predictive of occult lymph node metastasis, with AUCs of 0.648 and 0.608, respectively. However, the TEG parameters were not predictive of occult metastases (P > 0.05).

Fig. 5
figure 5

ROC curves of preoperative coagulation parameters for the prediction of occult lymph node metastasis. A-B, Conventional coagulation indices. cN0 OSCC patients (N = 241) were included in ROC curve analysis, in which 66 patients were confirmed to have lymph node metastases. AUC, area under curve. P < 0.05*, P < 0.01**, P < 0.001***

Full size image

Discussion

In this study, we innovatively elucidated the preoperative coagulation state in OSCC patients by combining TEG technology with conventional coagulation indices. We found that OSCC patients had an increased R time, K time, Angle and MA, indicating a whole-time-phase hypercoagulable state in the initial stage of coagulation, the early stage of blood clot formation, and the stage of coagulation cascade amplification. Though it is widely accepted for the use of TEG in numerous clinical scenarios such as surgery, trauma, and sepsis, its predictive value for disease progression in cancer patients has not been commonly recognized, with only studies involving relatively small case numbers reported. Zhou et al. revealed that preoperative TEG parameters such as K time, R time and MA can predict the clinical stage of lung cancer [12]. Another study revealed that Angle was closely related to postoperative local recurrence in pancreatic cancer patients, overall survival and disease-free survival [13]. Patients with papillary thyroid cancer had shorter R and K times, a higher Angle, and a higher CI than did the patients in the control group [14]. However, to our knowledge, in head and neck cancer, studies have been conducted only on the postoperative monitoring of microvascular anastomosis via TEG technology, but the oncological predictive value of TEG parameters has not been explored [15, 16]. Our study revealed for the first time that the predictive effect of TEG on OSCC tumour staging is good, further expanding its application value in the field of head and neck tumours. However, the potential impact of tobacco and areca nut use on coagulation function cannot be ruled out. Studies have confirmed that smoking may promote a hypercoagulable state [17, 18], while the effects of various components of areca nut on coagulation are still unclear [19, 20]. Further baseline matching with control groups is needed to support the hypercoagulable state in OSCC patients.

There was a significant correlation between tumour T stage and various TEG parameters, which indicated that the primary lesion was the culprit of the cancer-associated hypercoagulable state, in which coagulation factors, platelets and fibrinogen all play important roles. On the one hand, tumour-derived inflammatory factors such as interleukin-6 (IL-6) can stimulate FIB production and platelet elevation [21]. Tumour cells can directly affect coagulation and fibrinolysis by secreting a variety of procoagulant substances, such as tissue factor (TF), prothrombin activator inhibitor-1 (PAI-1), adenosine diphosphate (ADP), and exosomes [22]. On the other hand, as the tumour grows larger, the hypoxic microenvironment, necrosis, and endothelial dysfunction within the tumour further promote local hypercoagulation [23,24,25].

Notably, decreased R and K times and increased Angle were closely associated with tumour N stage, suggesting that lymph node metastasis influences mainly the early stages of coagulation. It has been observed that the fibrinogen α chain can pass through the integrin-AKT pathway to promote lung cancer growth and metastasis [26]. Another promising finding was that fibrin and platelets can adhere to circulating tumour cells (CTCs) to form microthrombi, which can not only help CTCs escape immune surveillance and shear forces in the blood but also enable their spread to distant organs and formation of metastases [27]. Our previous study demonstrated that intratumoural platelet microthrombi are an independent risk factor for lymph node metastasis in OSCC patients [28], suggesting that platelets are also important members of the tumour microenvironment. Recent studies have also reported a reduction in tumour lung metastases in mouse models in which platelet activation is inhibited, suggesting that platelets may be involved in the occurrence of metastasis [29, 30]. In TEG parameters, the K time can to some extent reflect early platelet activation. The decreased K time in N + patients suggest that platelet early activation may be involved in metastasis. The MA value also reflects platelet function and quantity; however, we found that it is primarily associated with T stage and does not significantly predict lymph node metastasis.

However, it is difficult to comprehensively and accurately evaluate the functional status of coagulation members alone through TEG. Our study also found that D-dimer and FIB are more sensitive in predicting lymph node metastasis. It is still necessary to further clarify how coagulation members affect tumour progression through high-throughput screening methods [31].

Conclusions

Patients with OSCC are in a hypercoagulable state, which tends to become more pronounced as the tumour develops. Tumour T stage affects the whole-time phases of coagulation whereas tumour N stage affects coagulation only in the early stages of coagulation. Preoperative TEG parameters in OSCC patients may serve as potential biomarkers for tumour evaluation.