Hematological Parameters as an Early Marker of Deep Vein Thrombosis in Diabetes Mellitus: An Observational Study

Background Although the association between the presence of diabetes mellitus and the development of deep vein thrombosis (DVT) is well known, the role of novel biomarkers in predicting the development of DVT in diabetic patients is not yet known to a large extent. Studies have shown that complete blood count (CBC) and CBC-derived parameters such as neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) can be used as surrogate markers to detect DVT. This study was conducted to assess the utility of NLR and PLR as a marker of DVT in diabetic patients. Methodology This case-control study was conducted among a calculated sample size of 109 DVT patients in one arm and 109 non-DVT patients in another arm. Hematological tests including total leucocyte count, differential leucocyte count, total neutrophil count, total lymphocyte count, total platelet count, NLR, and PLR were performed. Results We found a significant difference in NLR and PLR between the DVT and the non-DVT groups. In addition, we found that NLR and PLR were significantly higher in the diabetic group, indicating the presence of inflammation in association with diabetes mellitus. Analysis of the receiver operating characteristic curve showed that at a cut-off value of 2.83, NLR can detect DVT in diabetic patients with 67% sensitivity and 92% specificity. Similarly, PLR at a cut-off value of 131.46 can detect DVT in diabetic patients with 56% sensitivity and 90% specificity. Conclusions We conclude that NLR and PLR are novel inflammatory markers that can help in the early detection of DVT in diabetic patients.


Introduction
White blood cells are associated with acute and chronic inflammation. Neutrophilto-lymphocyte ratio (NLR) is an easily available marker for the evaluation of chronic inflammatory disorders [1]. It is derived from the total neutrophil and total lymphocyte counts of the complete blood count (CBC). NLR has been shown to have clinical relevance in metabolic syndrome [2], atherosclerosis, and cardiovascular disease [3]. The prevalence of diabetes mellitus in India is around 7.3% [4], and 14.9% of patients with diabetes mellitus develop deep vein thrombosis (DVT) during their lifetime [5]. DVT carries a grave prognosis in diabetes mellitus with one-year mortality of around 4.7-15.4% [6]. Post-thrombotic syndrome develops in 25-50% of patients with DVT within two years which seriously affects their quality of life [7].
DVT involves the activation of a network of coagulation factors, and the process is hastened by inflammatory mediators within the microenvironment. It is associated with impaired fibrinolysis [8]. Systemic inflammation has a modulatory role in thrombosis [9]. Hence, the CBC may provide an idea about the inflammatory status of the patient, which can indirectly indicate the procoagulant/prothrombotic state. From the CBC, various clues regarding the inflammatory status of the body can be obtained. CBC parameters that can provide a wide range of details regarding the type and duration of the inflammation are available 1 2 3 1 4 1 even at the grassroots levels of the healthcare system. They can be widely used and easily interpreted to detect various inflammatory conditions. The derived parameters such as NLR and platelet-to-lymphocyte ratio (PLR) are recently being used as new diagnostic criteria for various inflammatory disorders. Few studies have shown the association of these derived parameters with acute DVT [5]. DVT has been found to be associated with high monocyte counts [8]. In addition, these parameters have been shown to be associated with the vascular complications of diabetes mellitus [9,10]. To date, very few studies have been conducted among diabetes patients, particularly in association with vascular complications such as DVT, using the CBC and its derived parameters as a marker of inflammation. There are still ongoing controversies regarding the use of CBC-derived parameters as an early diagnostic marker of DVT or to assess the prognosis. Although some recent studies have shown an association of CBC parameters with DVT, it has not been utilized as a diagnostic marker. Similarly, recent studies have shown the association of CBC parameters with complicated diabetes mellitus [11], but they fail to answer the question of whether the increase in the inflammatory parameters is a result of the thrombotic microvascular complications or due to the diabetic status per se. Hence, there might be a correlation between the CBC-derived parameters and the diabetes status of an individual [12]. The use of CBC-derived parameters as an early predictor of DVT in diabetic patients is an area that has not been investigated and needs to be studied further. As 14.9% of diabetic patients may develop DVT, it is a burden on the healthcare facility which needs to be addressed. Hence, the objective of our study was to determine a suitable early marker for DVT in diabetic patients and assess the utility of NLR and PLR as a marker of DVT to reduce the associated mortality and morbidity with DVT in these patients.
This study aimed to investigate different hematological parameters in patients with DVT with and without diabetes mellitus. The primary objective was to evaluate the mean difference in NLR and PLR between patients with DVT and healthy controls. The secondary objective was to evaluate the difference in NLR and PLR in patients with DVT with and without diabetes mellitus.

Materials And Methods
This case-control study evaluated the changes in the total leukocyte count, NLR, and PLR in DVT patients attending the Department of Cardiovascular Surgery, All India Institute of Medical Sciences (AIIMS), Bhubaneswar, India. In addition, we compared the abovementioned parameters with age and gendermatched controls from AIIMS, Bhubaneswar, India. The sample size was calculated using a sample size calculation formula for case-control studies (https://communitymedicine4asses.files.wordpress.com/2021/10/sample-size-formula-for-case-controlstudies) with a power of 80% and an attrition rate of 15%. After obtaining approval from the Institutional Ethical Committee of AIIMS, Bhubaneswar (approval number: IEC/AIIMS BBSR/PG Thesis/2019-20/10), the study was conducted with 109 participants in each group between September 15, 2019, to December 15, 2020. Patients with hematologic diseases, active infections, malignancies, and cardiac, renal, or hepatic disorders were excluded from the study. Those who refused to give consent were also excluded from the study. The study initially enrolled 250 patients, 20 of whom were not willing to give consent for participation, and another 12 patients were excluded based on the exclusion criteria. Finally, 238 patients were enrolled in this study. They were divided into the following two groups: group A (DVT) and group B (non-DVT). Group A or the DVT arm consisted of 119 patients, and group B or the non-DVT arm consisted of 119 patients. Group A or the DVT arm was again divided into the following two subgroups: subgroup 1 with diabetes mellitus (n = 39), and subgroup 2 without diabetes mellitus (n = 70). Group B or the non-DVT arm was divided into the following two subgroups: subgroup 1 with diabetes mellitus (n = 52), and subgroup 2 without diabetes mellitus (n = 57).
All participants were interviewed using a self-made questionnaire. Their demographic data were collected, their medical history was assessed, and their medical reports were reviewed. Patients with fasting blood sugar levels of more than 126 mg/dL, postprandial blood sugar levels of more than 200 mg/dL, and HbA1c of more than 6.5% were grouped as diabetic [13,14]. Patients' weight (in kg), height (in m), heart rate (in beats per minute), and blood pressure (in mmHg) were recorded in the outpatient department for future reference.
All participants underwent blood sample collection by a finger prick method, and capillary blood was obtained. Total leukocyte counts were determined manually using Turk's dilution fluid in Neubauer's chamber. Total platelet counts were determined manually in Neubauer's chamber using an ammonium oxalate solution. Differential leukocyte counts were determined in a peripheral blood smear obtained by a finger prick and stained with Leishman stain. Using the percentage values obtained from the differential count and the total leukocyte counts, total neutrophil count, and total lymphocyte count were calculated as total neutrophil counts = % of neutrophils in differential count × total leukocyte count/100 cells/mm 3 , and total lymphocyte count = % of lymphocytes in differential count × total leukocyte count/100 cells/mm 3 . From these values, NLR and PLR were calculated.
The data were analyzed using R software (2022). The continuous variables were expressed as mean ± SD, and the categorical variables were expressed as median and interquartile range. All continuous data were tested for normality using the Shapiro-Wilk test. Normally distributed data in both the groups were tested by unpaired Student's t-test, and non-normally distributed data were tested using the Mann-Whitney test to evaluate the difference between the groups. Spearman's correlation analysis was performed to correlate two continuous variables. Receiver operator characteristic (ROC) curve analysis was performed to determine a cut-off value of NLR and PLR for detecting DVT. Sensitivity, specificity, and area under the curve (AUC) for delineation of cut-off points for NLR and PLR were also calculated.   Table 2 presents the comparison of variables among diabetic and non-diabetic patients having DVT. Diabetic patients having DVT were elderly, had higher total leucocyte count, total neutrophil count, total lymphocyte count, higher platelet count, higher monocyte count, NLR, PLR, and higher HbA1C (p < 0.05) than diabetic patients without DVT. Table 3 presents the comparison of variables among diabetic and non-diabetic patients without DVT. Diabetic patients without DVT had higher BMI, higher total leucocyte count, total neutrophil count, total lymphocyte count, platelet count, monocyte count, NLR, PLR, and higher HbA1C (p < 0.05) than non-diabetic patients without DVT.   The correlation between NLR and HbA1C was analyzed and the result revealed a positive correlation (r = 0.7361), and it was statistically significant (p = 0.001) (Figure 1). The correlation between NLR and BMI has been analyzed and the result revealed a positive correlation between the two (r = 0.7337), and it was statistically significant (p < 0.001) (Figure 2).

Discussion
The NLR value in the DVT group was 2.37, which was significantly higher than the NLR value of 1.98 in the control group. Elevated NLR is a better marker of enhanced inflammation in DVT patients because NLR takes into account the neutrophil count and the lymphocyte count, with neutrophils serving as the first line of defense and lymphocytes regulating inflammation. PLR was found to be significantly higher in the DVT group compared to the control group. PLR acts as a combined marker of platelet aggregation and inflammatory status as it takes into account two components, namely, platelets and lymphocytes. Our findings were similar to those reported by Ming et al. [5], Erden et al. [15], and Tural et al. [16]. Taking NLR as a parameter for ROC analysis, the AUC for NLR was 0.722, which indicates that an increase in NLR serves as a predictive marker of DVT. The sensitivity and specificity of NLR to detect DVT were found to be 63% and 65%, respectively, at a cut-  [15]. Using the above-mentioned cut-off values, we assessed the utility of the double positivity of NLR and PLR in predicting DVT, and it was found that the double positivity of both NLR and PLR was able to predict DVT with a sensitivity of 63% (53.5-72.3%) and a specificity of 67%. We also correlated NLR and PLR values with HbA1C to determine any correlation between the inflammatory status and long-standing glycemic control in diabetic patients. In our study, the NLR and PLR positively correlated with the levels of HbA1C, which is a marker of long-term blood sugar control. The positive association between NLR, as well as PLR with HbA1C, can be attributed to the increased levels of inflammation associated with uncontrolled blood glucose levels. Our study established that the novel markers NLR and PLR can predict DVT in diabetic individuals beyond values of 2.83 and 131.46, respectively. These simple novel biomarkers can predict DVT in diabetic patients with reliable sensitivity and specificity even in small peripheral hospitals.

Conclusions
NLR and PLR are novel biomarkers to detect DVT even in diabetic patients. Our study demonstrated that NLR and PLR were significantly (p < 0.05) higher in patients with DVT compared to patients without DVT. NLR and PLR were significantly higher among diabetic patients with DVT compared to non-diabetic patients with DVT. NLR as a diagnostic parameter has a sensitivity of 67% and specificity of 92% in detecting DVT among diabetic patients, and PLR as a diagnostic parameter has a sensitivity of 56% and specificity of 90% in detecting DVT among diabetic patients. These two novel biomarkers will add to the prediction and prevention of DVT among diabetic patients in the near future.

Additional Information Disclosures
Human subjects: Consent was obtained or waived by all participants in this study. Instiutional Ethics Committee (IEC), All India Institute of Medical Sciences (AIIMS), Bhubaneswar, Odisha, India issued approval IEC/AIIMS BBSR/PG Thesis/2019-20/10. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue. Conflicts of interest: In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.