Glycemic Control and Plasma Levels of Pro-Inflammatory and Pro-Thrombotic Biomarkers in Diabetic Patients Presenting with Acute Pulmonary Embolism

Abstract

Type I and type II diabetes are closely associated with a pro-inflammatory state and to a pro-thrombotic state. The role of glycemic control in pulmonary embolism (PE) is poorly understood and requires additional investigation. The aim of this study is to investigate the relationship between glycemic control and thrombo-inflammatory biomarkers in a PE patient cohort compared to normal samples. Demographic and clinical information for 86 diabetic patients and 106 non-diabetic patients presenting with acute PE was collected via retrospective chart review. Plasma levels of pro-inflammatory (C-reactive protein [CRP], tumor necrosis factor-alpha [TNF-α], interleukin-6 [IL-6]) and pro-thrombotic (d-dimer, plasminogen activator inhibitor-1 [PAI-1], tissue plasminogen activator [tPA], thrombin activatable fibrinolysis inhibitor [TAFI], von-Willebrand factor [vWF], endogenous glycosaminoglycans [GAGs]) biomarkers were drawn within 24 hours of diagnosis of acute PE. Data was also obtained for a population of healthy adult controls. All the pro-inflammatory and pro-thrombotic biomarkers were elevated in diabetic PE patients in comparison to healthy controls. None of the biomarkers were elevated in diabetic PE patients when compared to non-diabetic PE patients. There was no difference in the levels of the pro-inflammatory biomarkers according to glycemic control. The plasma level of TAFI was elevated in diabetic patients with poor glycemic control. Diabetic patients were more likely to have a more severe PE. These studies demonstrate that thrombo-inflammatory biomarkers are elevated in diabetic PE patients with associated comorbidities in comparison to normal individuals. However, there is no difference between the PE cohort alone in comparison to PE with diabetes. The role of TAFI within the continuum of diabetic vascular disease warrants additional investigation.

Keywords

pulmonary embolism diabetes mellitus HbA1c TAFI

Introduction

Venous thromboembolism (VTE) is estimated to affect between 350,000 and 600,000 individuals annually in the United States. The mortality rate is approximately 100,000 people each year.¹ Diabetes mellitus (DM) is one of the most common metabolic disorders and is characterized by longstanding hyperglycemia and insulin resistance. Complications of longstanding DM often result in end-organ dysfunction and failure in the kidneys, nerves, retina, heart, and blood vessels. The prevalence of DM is rapidly increasing from 1990 to 2017 the global disease burden is estimated to have increased from 11.3 million to 22.9 million, while the global prevalence of diabetes increased from 211.2 million to 476.2 million in that period.² This increased disease burden has been observed in both developed and underdeveloped countries and is driven primarily by an increase in the incidence of type two diabetes.³ This increase is also associated with massive increases in healthcare spending both for individual patients and for healthcare systems around the world.

Both type I and type II DM are well established and important risk factors for cardiovascular disease and its subsequent microvascular and macrovascular complications. Cardiovascular disease is the leading cause of both morbidity and mortality among patients with diabetes and often presents ten to twenty years earlier in diabetic patients.^4,5 The most common cardiovascular complications experienced by diabetic patients include atherosclerosis, myocardial infarction, and stroke. The continuum of vascular disease experienced by diabetic patients results in long-term stress on the vascular endothelium and the sequelae associated with long-term vascular strain. Vascular damage is a recognized and foundational driver of hypercoagulability, and a pro-thrombotic state has come to be viewed as an important component of diabetes and the metabolic syndrome. So closely tied is DM to thrombosis that it is a component of the CHA₂DS₂-VAS_c score that assesses the risk of cardioembolic events in patients with atrial fibrillation. Early work attributed this pro-thrombotic state to increased fibrinogen and plasminogen activator inhibitor-1 (PAI-1) as well as other platelet abnormalities, and since then a variety of markers of endothelial dysregulation have been found to be elevated in diabetic vascular disease.^6,11 Work to understand the exact mechanisms behind this pro-thrombotic state is ongoing. There are several different factors that likely contribute to thrombotic events in DM; these include longstanding inflammatory states, primary hemostatic changes, increased levels of various clotting factors, impaired fibrinolysis, increased oxidative stress, and endothelial dysfunction.^6–12 Insulin resistance is itself a determinant of endothelial dysfunction as insulin signaling pathways are critical to endothelial generation of nitric oxide (NO), endothelin-1, and adhesion molecules. Individuals with prediabetes, type II DM, and newly diagnosed type I DM have all been demonstrated to have biochemical indications of endothelial dysfunction reflected by high levels of circulating tissue plasminogen activator (tPA) and von-Willebrand factor (vWF).¹³ These changes to the coagulation and fibrinolytic systems often translate to real-world impacts on patients—some studies have demonstrated a nearly two-fold increase in the risk for deep vein thrombosis and pulmonary embolism (PE) in patients with diabetes.¹⁴

Evidence is accumulating that both type I and type II DM are closely related to underlying chronic low-grade inflammatory states.⁹ Extensive prior work has demonstrated a positive correlation between patients with DM and increased plasma measurements of common systemic inflammatory biomarkers and cytokines involved in the acute phase response including tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), and C-reactive protein (CRP).^9–11 A significant driver of this pro-inflammatory state in type II diabetics is thought to be obesity, specifically the hypoxic changes that occur in adipose tissue as it expands in mass disproportionate to the development of the surrounding capillary network.^12,15 In a different vein, the hallmark of type I diabetes is chronic inflammation resulting in reduced or absent function of insulin producing beta cells in the pancreatic islets of Langerhans. Some acute phase markers, particularly TNF-α, have also been directly linked to the development of insulin resistance via inflammation of pancreatic islet beta cells.¹⁶ Glucolipotoxicity is a recognized driver of the production of reactive oxygen species with downstream damage to the endothelium in both type I and type II diabetes. Chronically elevated glycemic load results in oxidative stress which directly contributes to the production of pro-inflammatory mediators.¹⁵ This state of chronic low-grade inflammation is tied closely to further pancreatic damage and insulin resistance.⁹ Large cohort studies have also closely linked higher levels of baseline CRP and IL-6 with development of the metabolic syndrome, with CRP widely accepted as the most sensitive overall marker.^17,18 It is becoming increasingly important to better understand how chronic activity of these acute phase components impacts the development and course of metabolic disorders including both type I and type II DM.

There is contrasting evidence that the degree of glycemic control in diabetic patients may impact these pro-thrombotic and pro-inflammatory states. In an analysis of 1018 adults with DM increasing hemoglobin-A1c (HbA1c) levels, a common maker of glycemic control or lack thereof, were closely associated with elevated CRP levels.¹⁹ Likewise, intensive glycemic control has been shown to be directly correlated with a reduction in highly sensitive CRP (hs-CRP).²⁰ Fibrinogen, another reactant of the acute phase that is closely related to thrombosis, has also been shown to be independently associated with HbA1c.²¹ Interestingly, a study examining 500 adults with recent onset type II DM found that initiation of insulin or metformin did not result in a reduction in hs-CRP or IL-6 levels despite patients achieving expected reductions in both blood glucose levels and HbA1c.²² It has been demonstrated that PAI-1 is elevated in both diabetic patients and patients with other conditions resulting in insulin resistance and hyperglycemia.²³ In summary, it is generally accepted that both type I and type II diabetes are closely associated with a chronic systemic pro-inflammatory state, and that these patients experience long-term stressors on the vascular endothelium which likely contributes to the development of a pro-thrombotic state. However, the exact role of glycemic control in this disease process is unclear and requires additional investigation.

Very few studies have attempted to explain the complex relationship between glycemic control in DM and the development of pro-thrombotic and pro-inflammatory states in atherosclerosis, endothelial damage, and eventually VTE and PE. HbA1c, when used as an indicator for glycemic control, may be an important measurement to quantify the level of thrombo-inflammatory and systemic inflammatory response that is observed in diabetic patients. The possible role of glycemic control in the levels of the thrombo-inflammatory biomarkers d-dimer, PAI-1, tPA, thrombin activatable fibrinolysis inhibitor (TAFI), vWF and endogenous glycosaminoglycans (GAGs), the acute phase protein CRP, and the systemic inflammatory cytokines IL-6 and TNF-α in diabetic patients acutely presenting with PE is not fully understood. Our study attempts to better characterize the complex relationship between glucose regulation in diabetic patients who have experienced PE and the levels of these circulating systemic and pro-thrombotic inflammatory biomarkers.

Materials and Methods

Plasma samples from 265 patients presenting with acute PE between January 2017 and October 2021 were collected following the ongoing institutional review board (IRB) and pulmonary embolism response team (PERT) protocols at Loyola University Medical Center and affiliated hospitals. PE was diagnosed by computed tomographic angiography or V/Q scintigraphy per admitting physician discretion. One hundred and thirteen of these patients had a concurrent diagnosis of type I or type II DM at the time of PE. Patients were considered to have active diabetes if this diagnosis was documented in their medical record or if they were receiving insulin or oral hypoglycemic agents at the time of PE.

The demographic and clinical information of these 113 diabetic PE patients, including age, sex, and body mass index (BMI) was collected via retrospective chart review of their electronic medical records. We also collected data regarding comorbid hypertension (HTN), congestive heart failure, coronary artery disease (CAD), peripheral artery disease, chronic obstructive pulmonary disease, chronic kidney disease (CKD), and their obstetric and cancer history if present. Medication usage and overall and 30-day mortality at the time of PE diagnosis were also obtained. Ninety-nine of these patients had a hemoglobin A1c (HbA1c) measurement collected within 3 months of their presentation with acute PE. We included patients who received an anti-diabetic treatment (n:49) or statin therapy (n:45) at the time of presentation to the hospital with PE. After excluding 15 patients with active cancer the 84 diabetic PE patients for whom a recent HbA1c was available were selected for final analysis.

Plasma samples were obtained from all of these patients during their hospitalization. Whole blood samples were drawn from patients within 24 hours of confirmed diagnosis of acute PE and collected under an IRB-approved protocol. Samples were collected in 3.8% (0.109 mol/l) sodium citrate tubes at the time of PE diagnosis, processed for platelet-poor plasma, and stored at −70°C prior to analysis. The levels of the acute phase protein (CRP), acute-phase cytokines (IL-6, TNF-α), and thrombo-inflammatory markers (d-dimer, PAI-1, tPA, TAFI, and vWF) were quantified using commercially available ELISA immunoassays for each of the PE and control plasma samples. The levels of endogenous GAGs were quantified by utilizing Heparin Red Probe (Redprobes UG, Munster, Germany). Adequate glycemic control was defined as HbA1c ≤ 7.0 in concordance with the most recently published HbA1c targets from the American College of Physicians.²⁴ Circulating levels of each biomarker and cytokine in diabetic PE patient plasma were compared in patients with good glycemic control (HbA1c ≤ 7.0 mg/dl, n:34) and poor glycemic control (HbA1c > 7.0 mg/dl, n:50).

The levels of each of these biomarkers and cytokines were also compared to plasma samples from PE patients who did not have DM (“non-DM”). Patients presenting to our hospital with PE and without diabetes also had plasma samples drawn within the same timeframe (January 2017 and October 2021) and levels of each biomarker were assessed in the manner described above. There were plasma samples obtained for 152 non-diabetic patients presenting with PE within the study period. After excluding patients with active cancer in this cohort, there were 106 non-diabetic patient plasma samples available which were compared for levels of each tested biomarker in the diabetic PE cohort. Additionally, control plasma samples (n = 15) from healthy, non-smoking, non-obese adults, aged 19 to 53, were purchased from a commercially available source (George King Biomedical, Overland Park, Kansas) (“Healthy Controls”). The plasma levels of each marker in the healthy control population were also compared to those in the diabetic PE cohort.

Statistical analysis was conducted using GraphPad Prism and Microsoft Excel software. Categorical variables were compared using the Chi-squared and Fisher's exact tests. The continuous variables and the statistical difference between PE groups and normal controls were evaluated utilizing nonparametric Mann–Whitney U, Student’s t-tests and Kruskal–Wallis ANOVA. A P value of less than .05 was considered significant.

Results

The mean age of diabetic PE patients in our study was 65.5 (± 14.1) years. There were 47 males (55.95%) and 37 females (44.05%) in the diabetic PE cohort. The average BMI of patients in the diabetic PE cohort was 33.51. Forty-seven of these patients were nonsmokers, 37 were former smokers, and 6 were current smokers. Four patients in this cohort were COVID-19 positive at the time of hospitalization. Within the diabetic PE cohort, the average HbA1c was 7.66; 50 patients had an A1c greater than 7 and 34 had an A1c less than 7. The mean age of non-diabetic PE patients was 59.25 (± 15.5). There were 49 males (46.67%) and 57 females (53.33%) in the non-diabetic PE cohort. The average BMI of patients in this cohort was 33.35. Seventy-one of these patients were nonsmokers, 24 were former smokers, and 11 were current smokers. None of these patients were reported as COVID-19 positive at the time of hospitalization. Demographic data was not available for the commercially obtained population of healthy controls. Regarding medical comorbidities the patients in the diabetic PE cohort had higher rates of cardiovascular disease (HTN, heart failure, CAD), chronic obstructive pulmonary disease, and CKD than patients in the non-diabetic PE cohort. Additional information regarding medical comorbidities, mortality, and PE severity is found in Table 1.

Table 1.

Demographic Information, Medical Comorbidities, and Mortality for the Diabetic Pulmonary Embolism and Non-Diabetic Pulmonary Embolism Cohorts.

Parameter	Diabetic Patients (n: 84)	Non-Diabetic Patients (n: 106)	P-Value
Age (mean, years)	65.5	59.25	<.01
Gender, n (%)	Male: 47 (55.95%)	Male: 49 (46.23%)	.18
Gender, n (%)	Female: 37 (44.05%)	Female: 57 (53.77%)	.18
BMI (Mean, kg/m²)	33.51	33.35	.95
HbA1c (Mean, %)	7.66	NA	NA
HbA1c < 7.0	34 (40.48%)	NA	NA
HbA1c ≥ 7.0	50 (59.52%)	NA	NA
Troponin I (Mean, ng/ml)	0.61	0.99	.03
Troponin I ≥ 0.04, n (%)	38 (45.24%)	48 (45.28%)	.99
BNP (Mean, pg/ml)	318.21	364.44	.16
NT-ProBNP (Mean, pg/ml)	1032.50	1187.35	.99
Lactate (Mean, mmol/l)	2.35	2.13	<.01
PE Severity, n (%)	Low risk: 29 (34.50%)	Low risk: 35 (33.00%)	.78
	Sub-massive: 49 (58.30%)	Sub-massive: 63 (59.40%)
	Massive: 6 (7.20%)	Massive: 8 (7.60%)
PESI Score (Mean)	114.49	102.63	.04
SPESI Score (Mean)	1.57	1.37	.14
Smoking Status, n (%)	Never smoker: 47 (55.95%)	Never smoker: 71 (66.98%)	.12
	Current smoker: 6 (7.14%)	Current smoker: 11 (10.38%)	.43
	Former smoker: 37 (44.5%)	Former smoker: 24 (22.64%)	<.01
DVT at the time of PE diagnosis n (%)	52 (61.90%)	48 (45.28%)	.02
Hypertension, n (%)	71 (84.52%)	45 (42.45%)	<.0001
Chronic Heart Failure, n (%)	26 (30.95%)	15 (14.15%)	<.01
Coronary Artery Disease, n (%)	28 (33.33%)	9 (8.49%)	<.0001
COPD, n (%)	11 (13.10%)	9 (8.49%)	.30
CVA, n (%)	22 (26.19%)	6 (5.66%)	.0001
Chronic Kidney Disease, n (%)	70 (83.33%)	7 (6.60%)	<.0001
Lowest SaO2	91.85	91.55	.63
Mortality, n (%)	7 (8.33%)	16 (15.09%)	.15
30-day mortality, n (%)	4 (4.76%)	6 (5.66%)	.78
Major Bleeding, n (%)	1 (1.19%)	2 (1.89%)	.70
VTE Recurrence, n (%)	13 (15.48%)	4 (3.77%)	<.01

As shown in Table 2, the levels of all of the tested pro-inflammatory biomarkers were elevated in diabetic PE patients in comparison to healthy controls (P < .05). CRP (43.11-fold) and IL-6 (73.83-fold) demonstrated the most prominent increases in comparison to healthy controls. The levels of all the tested pro-thrombotic biomarkers were likewise elevated in diabetic PE patients in comparison to healthy controls (P < .05). d-Dimer (29.78-fold) and PAI-1 (9.67-fold), in addition to endogenous GAGs (3.56-fold), demonstrated the largest increase in comparison to healthy controls. None of the measured pro-inflammatory and pro-thrombotic biomarkers in diabetic PE patients were significantly elevated when compared to non-diabetic PE patients (Table 3).

Table 2.

Plasma Levels of Pro-Inflammatory and Pro-Thrombotic Biomarkers Diabetic PE Patients Compared to Healthy Controls.

Biomarker	Diabetic (n = 84)	Healthy Controls (n = 15)	Fold Increase	P-Value
(A) The Levels of Systemic Inflammatory Biomarkers in Comparison to Healthy Controls (Mean, IQR)
CRP (mg/l)	18.11 (12.45-16.98)	0.42 (0.00-0.52)	43.11	<.0001
IL-6 (pg/ml)	66.13 (11.75-60.28)	0.90 (0.62-1.09)	73.83	<.0001
TNF-A (pg/ml)	1.46 (0.77-1.63)	1.36 (1.38-1.77)	1.06	.04
(B) The Levels of Thrombo-Inflammatory Biomarkers in Comparison to Healthy Controls (Mean, IQR)
d-Dimer (ng/ml)	6238.81 (2860.49-9134.32)	209.44 (121.71-289.05)	29.78	<.0001
PAI-1 (ng/ml)	54.81 (33.96-67.47)	5.67 (3.24-8.24)	9.67	<.0001
t-PA (ng/ml)	11.83 (6.44-16.48)	3.20 (2.18-3.84)	3.47	<.0001
TAFIa (%)	99.93 (91.02-110.11)	87.94 (78.18-98.91)	1.13	.01
VWF (%)	138.16 (130.05-145.82)	76.14 (65.57-84.21)	1.81	<.0001
Endogenous GAGs (μg/ml)	2.14 (0.82-2.50)	0.60 (0.41-0.77)	3.56	<.0001

Table 3.

Plasma Levels of Pro-Inflammatory and Pro-Thrombotic Biomarkers Diabetic PE Patients Compared to Non-Diabetic PE Patients.

Biomarker	Diabetic (n = 84)	Non-Diabetic (n = 106)	Fold Increase	P-Value
(A) The Levels of Systemic Inflammatory Biomarkers in Comparison to Non-Diabetic Patients (Mean, IQR)
CRP (mg/l)	18.11 (12.45-16.98)	22.12 (11.51-16.46)	0.81	.55
IL-6 (pg/ml)	66.13 (11.75-60.28)	76.47 (6.66-61.34)	0.86	.81
TNF-A (pg/ml)	1.46 (0.77-1.63)	1.50 (0.88-1.69)	0.97	.59
(B) The Levels of Thrombo-inflammatory Biomarkers in Comparison to Non-Diabetic Patients (Mean, IQR)
d-Dimer (ng/ml)	6238.81 (2860.49-9134.32)	6856.66 (2827.91-10295.57)	0.90	.44
PAI-1 (ng/ml)	54.81 (33.96-67.47)	51.74 (35.31-67.41)	1.05	.55
t-PA (ng/ml)	11.83 (6.44-16.48)	13.28 (8.22-18.21)	0.89	.21
TAFIa (%)	99.93 (91.02-110.11)	97.44 (91.07-103.43)	1.02	.44
VWF (%)	138.16 (130.05-145.82)	142.51 (128.74-145.75)	0.96	.40
Endogenous GAGs (μg/ml)	2.14 (0.82-2.50)	1.87 (0.86-2.23)	1.14	.62

As shown in Table 4, there was no significant difference in the levels of the tested systemic inflammatory biomarkers and cytokines in diabetic PE patients with adequate glycemic control when compared to diabetic PE patients with poor glycemic control. The mean CRP level in patients with HbA1c ≤ 7.0 mg/dl was 14.93 μg/ml; the mean CRP level in patients with HbA1c > 7.0 mg/dl was 17.69 μg/ml (P = .83). The mean level of IL-6 in patients with adequate glycemic control was 75.93 pg/ml, the mean level was decreased to 58.57 pg/ml in patients with poor glycemic control (P = .65). When assessing levels of TNF-α there was a mean plasma concentration of 1.46 pg/ml in patients with adequate glycemic control and a likewise mean plasma concentration of 1.46 pg/ml in patients with poor glycemic control (P = .47).

Table 4.

Plasma Levels of Pro-Inflammatory and Pro-Thrombotic Biomarkers Diabetic PE Patients Stratified According to Glycemic Control.

Biomarker	HbA1c < 7.0 (n = 34)	HbA1c ≥ 7.0 (n = 50)	P-Value
(A) The Levels of Systemic Inflammatory Biomarkers According to HbA1c Levels (Mean, IQR)
CRP (mg/l)	14.93 (13.13-15.97)	17.69 (12.18-16.23)	.83
IL-6 (pg/ml)	75.93 (11.29-68.85)	58.57 (12.79-44.30)	.65
TNF-A (pg/ml)	1.46 (0.78-1.42)	1.46 (0.77-1.79)	.47
(B) The Levels of Thrombo-inflammatory Biomarkers According to HbA1c Levels (Mean, IQR)
d-Dimer (ng/ml)	7169.30 (3805.75-10917.19)	5761.59 (2553.94-8605.00)	.21
PAI-1 (ng/ml)	59.46 (3.52-8.42)	52.89 (32.37-64.39)	.28
t-PA (ng/ml)	11.57 (1.68-2.94)	11.97 (6.44-16.12)	.99
TAFIa (%)	91.46 (88.03-101.17)	104.27 (92.72-113.13)	<.01
VWF (%)	131.86 (126.25-141.82)	141.31 (134.79-152.31)	.07
Endogenous GAGs (μg/ml)	2.11 (0.84-2.45)	2.15 (0.82-2.50)	.96

The plasma level of TAFI was significantly elevated from 91.46% in diabetic patients with adequate glycemic control to 104.27% in patients with poor glycemic control (P < .01). The plasma level of vWF in patients with poor glycemic control was 131.86; the plasma level of vWF in patients with adequate glycemic control was 141.31. The difference in plasma vWF levels between these patient cohorts approached but did not reach significance (P = .07) and was significant when tested before the exclusion of patients with active cancer. We did not observe any significant difference in the levels of d-dimer, CRP, PAI-1, t-PA, and endogenous GAGs according to HbA1c levels within the diabetic PE cohort.

Discussion

The persistently increasing morbidity of DM and its resulting impact on individual patients and healthcare systems indicates that it is critical to continue to explore both the etiology of this disease process and the sources of potential mortality that patients may be at increased risk for. It is also important to consider the role glycemic control plays in this picture—though access to care remains non-universal, many diabetic patients will at some point begin treatment for this disease with lifestyle modifications or with a variety of glucose regulating medications. Additional questions regarding treatment choice, adherence and their impacts on glycemic control are out of the scope of this study; we sought to better understand how degree of glycemic control at the time of acute PE may impact plasma levels of a variety of pro-inflammatory and pro-thrombotic biomarkers in diabetic patients as well as if there was any correlation with PE severity and mortality. Our interest was in PE specifically, rather than other thrombotic events, because prior studies have closely tied DM with increased rates of PE and because it is an area of specific clinical interest for several physicians at our medical center.²⁵ To the best of our knowledge there are very few papers investigating pro-inflammatory and procoagulant responses in diabetic patients presenting with PE within the context of the level of glycemic control.

It is becoming widely accepted that the pathophysiology of DM involves a chronic inflammatory response that may be present even before the disease is diagnosed, and that the vascular strain associated with this response results blood vessel dysfunction and damage. In this sense, diabetes is closely tied to vascular insult, a key component of the triad of thromboembolic risk factors. As such, we elected to study a variety of markers central to the acute phase response and to two components of thrombosis: endothelial damage and fibrinolysis. To assess systemic inflammation in our patient population we measured plasma levels of TNF-α, IL-6, and CRP. TNF-α is central to the innate immune response and has a role in promoting cytokine production, expression of cellular adhesion molecules, and stimulation of cell growth and proliferation as well as upregulation of the major histocompatibility complex I and II molecules. Notably, TNF-α is closely tied with upregulation of major histocompatibility complex II on the vascular endothelium and transmigration of inflammatory cells through the vasculature. IL-6 is expressed by a variety of immune cells during the acute phase response and increased levels of IL-6 have been associated with chronic inflammation. CRP is an acute phase protein that is recognized as the most sensitive marker of the acute phase response that rises and falls rapidly as the disease process progresses.²⁶

To assess the presence or absence of a pro-thrombotic state in our patient population we measured plasma levels of d-dimer, PAI-1, tPA, TAFI, vWF, and endogenous GAGs. CRP was also assessed as a marker of a thrombotic state as it has been associated with an impaired fibrinolytic response.²⁷ d-Dimer is a common marker of thrombosis and tPA is a primary marker of the fibrinolytic system. PAI-1 is an inhibitor of the plasminogen system and upregulation of this biomarker has been related to fibrinolytic deficit in diabetic patients.²⁸ TAFI plays a central role in blood clot stabilization. CRP, d-dimer, tPA, PAI-1, and TAFI are commonly used to assess the activity or impairment of the fibrinolytic system. The vWF is a glycoprotein secreted by endothelial cells that is involved in thrombus formation.²⁹ Endogenous GAGs are components of the endothelial glycocalyx that regulate the access of cells and other bloodstream molecules to the endothelium. Severe inflammation can cause glycocalyx dysfunction resulting in matrix metalloproteinases dissolving portions of the glycocalyx and releasing endogenous GAGs into the vascular circulation. Though their relevance has not yet been assessed in large scale studies it is likely that these may also serve as a key method of quantifying endothelial dysfunction.³⁰ Together with vWF, we considered these to be markers of endothelial dysfunction.

In our study, all the tested pro-inflammatory and pro-thrombotic biomarkers and cytokines were significantly elevated in patients from our medical center who presented with a PE when compared to a population of commercially obtained controls. We expected this result, as the plasma samples from the healthy control population were meant to reflect a patient with none or few medical comorbidities or acute problems that may drive elevation of such biomarkers. Interestingly none of the tested biomarkers were significantly elevated when diabetic patients who presented with PE were compared to non-diabetic patients who presented with PE. We found this observation to be unexpected and inconsistent with the current literature regarding these biomarkers in diabetic patients.^6,9,11,13 This result was compounded by the significantly increased frequency of atherosclerotic disease states (HTN, CAD, CKD) that were observed in the diabetic PE cohort. In a cohort of patients with diabetes, especially one with a significantly increased frequency of atherosclerotic sequelae that are also associated with inflammatory responses, we expect to see an elevation in most if not all of these biomarkers in comparison to the non-diabetic PE cohort. There are several reasons why we may have been unable to observe this. It is possible that the lack of difference here is a reflection of the relatively sick patient cohort which we studied, with a number of patients in both the diabetic PE and non-diabetic PE cohorts suffering from a variety of comorbid conditions that may drive elevations of these inflammatory markers to various degrees including cardiovascular disease, pulmonary disease, and renal dysfunction. Because our study focused on patients presenting to the hospital with an acute PE, it was difficult to exclude patients with other significant comorbidities and each patient may or may not have been using medications such as anti-inflammatory drugs or statins at the time of admission that could potentially further alter the plasma level of each biomarker that we assessed.

The question that we were most interested in answering in this study was whether or not the degree of glycemic control in diabetic patients had any impact on both the pro-inflammatory and pro-thrombotic state that is a hallmark of DM. The current body of work on the topic of thrombogenic potential according to level of glycemic control in diabetic patients is inconclusive. There are established links between the mechanisms of hyperglycemia and thromboembolism. Most notably, advanced glycation end products contribute to endothelial dysfunction, as well as to oxidative stress of the vascular endothelium which has been demonstrated to increase tissue factor production.^31,32 However, it has been difficult to connect these endothelial insults with real world impacts on patients within the context of glycemic control. Some studies have demonstrated increased risk of VTE in female patients with HbA1c > 7.0%, while others have had difficulty demonstrating meaningful increased risk of VTE events in diabetics with poor glucose control.^33–35 Data focusing specifically on the risk of embolism within the context of glycemic control is also uncommon. One study by Fangen et al, noted an increased risk of embolic stroke in Dutch patients with atrial fibrillation and poor glycemic control when compared to patients with atrial fibrillation and good glycemic control. However, this study differs from ours in that all patients suffered from atrial fibrillation, itself a potential driver of clot formation and rupture. This study also included patients with an HbA1c available within a 2-year time period, rather than the more conventional 3-month timeframe that we elected to assess.³⁶

To our knowledge very few studies to date have focused on specifically answering this question, especially by measuring common biomarkers of inflammation and pro-coagulability. We hypothesized that poor glycemic control may lead to increased vascular strain and endothelial damage and that this could be reflected by changes in plasma levels of the pro-inflammatory and pro-thrombotic biomarkers discussed above. Within our cohort of diabetic PE patients our findings suggest that there is no direct association between degree of glycemic control and the levels of the inflammatory biomarker CRP and the inflammatory cytokines IL-6 and TNF-α in diabetic patients who have had a PE. Interestingly enough, this reinforces findings from Pradhan et al, that found no reduction in plasma levels of hs-CRP or IL-6 in diabetic patients who were started on insulin or metformin to achieve or improve their level of glycemic control.²¹ Our findings also possibly corroborate those of Isordia-Salas et al, who found plasma levels of CRP and fibrinogen in diabetic patients to be predominantly determined by BMI rather than by degree of glycemic control.³⁷ While the consensus within the literature is that a diabetic patient is experiencing a constant low-grade state of chronic inflammation, our results highlight that achieving improved control of diabetes may not directly reduce circulating plasma levels of these common inflammatory markers.

Our findings support also that the increased levels of thrombo-inflammatory biomarkers and dysregulation of hemostatic pathways may be important pathophysiological mechanisms in diabetic PE patients. Increased levels of TAFI, and potentially vWF, may be the potential contributors of impaired fibrinolysis and endothelial dysregulation in diabetic PE patients who have poor glycemic control. Work on the impact of glycemic control on thromboembolic events and in the quantification of plasma biomarkers of thrombosis in diabetic patients is ongoing. A 1994 article from Steiner et al notes increased risk of thrombosis and elevated plasma levels of vWF, E-selectin, and vascular cell adhesion molecule in diabetic patients. In this study the plasma level of each of these biomarkers, including vWF, was found to be elevated in diabetics independently of glycemic control.³⁸ Studies from Rosove et al and Knobl et al likewise found no decrease in several other pro-coagulant biomarkers, including vWF, in diabetic patients who achieved improved glycemic control.^39,40 More recently a study that sought to assess blood thrombogenicity within the context of glycemic control produced extremely interesting results. Osende et al were able to demonstrate that improving glycemic control in a patient with type II DM does in fact reduce overall blood thrombogenic potential, assessed by testing samples of blood from diabetic patients in an ex vivo Badimon perfusion chamber. They found a significant reduction in overall blood thrombus formation within those patients who were able to achieve improved glycemic control during the study timeframe, but notably did not observe significant decreases in in several pro-thrombotic biomarkers that they tested, including PAI-1.⁴¹ However, it should be noted that in their study the average level of PAI-1 was not elevated from a healthy baseline in their diabetic patient cohort as it was in ours. In summary, many of the results from our study reinforce those currently available—diabetic patients often have elevated plasma levels of several pro-inflammatory and pro-thrombotic biomarkers when compared to healthy individuals, but that degree of glycemic control may not have a major impact on plasma levels many of these markers of inflammation, thrombosis, and hemostasis. We believe that the role of TAFI specifically within the continuum of diabetic vascular disease is an underexplored area that warrants further investigation.

Another interesting finding from our study is that the mean PESI score was elevated in the diabetic PE cohort when compared to the non-diabetic PE cohort. The PESI score is a metric that seeks to classify PE severity by incorporating a patient's age, sex, and medical comorbidities including history of cancer, heart failure, chronic lung disease, and various vital sign measurements at the time of PE. This criterion does not specifically include diabetes. Patients are then classified into various tiers that are associated with rates of 30-day mortality. Type II DM has been tied to increased risk of in-hospital mortality in patients suffering from PE, and while not the focal question of our study the mortality findings we gathered did not corroborate this.^42,43 However, our results do agree with other studies that have tied preexisting diabetes with increased severity of PE.⁴⁴

Our study also has several limitations which should be noted. First, this is a retrospective study. We did not focus solely on type I or type II diabetic patients as the primary aim of this study was to assess levels of each biomarker in insulin resistant or dependent patients according to glycemic control or lack thereof. As many of the patients in this study presented to the hospital while taking other medications for their various comorbidities, we were not able to exclude those patients taking medications which may impact the plasma levels of some biomarkers assessed, including anti-inflammatory medications and cholesterol-lowering medications. The retrospective nature of this study also meant that it was difficult to exclude other drivers of thrombosis from our patient cohort, most particularly prolonged immobilization which may be a contributor especially in elderly patients, but which is difficult to assess from medical record review. This retrospective study included 106 non-diabetic PE patients and 84 diabetic patients with PE providing us two cohorts to demonstrate the differences in the biomarker levels in order to determine the role of glycemic index elevation on each respective level. Since this was a pilot study our observation clearly suggests the need for a larger study with an increased number of both groups to validate our data. The relatively small number of normal samples was included as a reference point and additional trials including a larger number of normal individuals may be needed, however, the fluctuations in glycemic index in the normal cohorts are not as obvious as in the patients with PE and diabetes as a comorbidity. Despite these limitations and the apparent lower number of representative patients in the PE group and normal individuals this study provides some observational differences in the levels if biomarkers in the three groups which require additional expanded studies. Finally, it has been suggested that additional genetic factors may also contribute in various ways to the plasma levels of all the assessed biomarkers and cytokines in this study, and these are factors for which we were not able to control.

Our study has several strengths. All of the patients assessed presented to and were evaluated at the same medical center within a 4-year window. Workup for each patient followed the Loyola University Medical Center PERT team protocols, and each plasma sample collected was stored and measured in the same fashion. We were able to exclude patients with active cancer, a significant driver of thrombosis, from our study. Moreover, the PERT program is still ongoing and recruiting additional patients and provides a platform to expand the current profiling of the biomarkers to better determine their diagnostic and prognostic values.

Summary and Conclusion

The low-grade chronic inflammation associated with DM likely contributes to endothelial dysregulation and may increase these patients’ predisposition to thromboembolic events including PE. Our findings suggest that the increased levels of thrombo-inflammatory biomarkers and dysregulation of hemostatic pathways are important pathophysiological mechanisms in diabetic PE patients. Increased levels of TAFI, and potentially vWF, may be the potential contributors of impaired fibrinolysis and endothelial dysregulation in diabetic PE patients who have poor glycemic control. Similarly, our findings reinforce earlier work that there is no direct association between the degree of glycemic control and the levels of the inflammatory biomarker CRP and the inflammatory cytokines IL-6 and TNF-α in diabetic patients acutely presenting with PE. In our study patients with diabetes who had PE were more likely to have a more severe PE however they were not at higher risk for overall and 30-day mortality. The clinical relevance of our study requires additional work to fully assess. We believe that all diabetic patients should be provided with the necessary resources to improve their glycemic control and to reduce their risk of associated microvascular complications when able. From our results, and from the results of those studies that we reviewed, the impact of seeking to improve glycemic control on a diabetic patient's risk of pulmonary embolus or other thromboembolic events remains unclear, however it is possible that diabetic control may contribute to less severe PE.

Further investigations are required to understand the complex relationship between these common markers of inflammation, coagulability, and the degree of glucose regulation in diabetic patients. The impact of BMI on plasma levels of these biomarkers also warrants additional investigation. Our results highlight the importance of glycemic regulation in diabetic PE patients and provide several interesting questions to investigate moving forward. Finally, prospective studies are needed to better determine if poor glycemic control directly contributes to the underlying inflammatory state experienced by diabetic patients, or if characteristics of this inflammatory state could be the drivers of poor glycemic control in some patients.

Footnotes

Acknowledgements

The authors are thankful to the staff of the clinical laboratories for the facilitation of the collection of blood samples. They are thankful to Dr Seth Robia and Dr Alain Heroux for their support. They are also grateful to Dr Lowell Steen Director of the Section of Cardiology, Department of Medicine and Dr Eva Vozjck Chair of the Department of Pathology and Laboratory Medicine for providing helpful advice and encouragement to carry out this research program. This study is partially funded by the Cardiovascular Research Institute, Loyola University Chicago. The skillful assistance of Ms. Erin Healy-Erickson in the preparation of this manuscript is gratefully acknowledged.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Jacob Pozin

Fakiha Siddiqui

Debra Hoppensteadt

Jeanine Walenga

Jawed Fareed

Bulent Kantarcioglu

References

Office of the Surgeon General (US); National Heart, Lung, and Blood Institute (US). The surgeon general’s call to action to prevent deep vein thrombosis and pulmonary embolism. Office of the Surgeon General (US); 2008.

Lin

Pan

, et al. Global, regional, and national burden and trend of diabetes in 195 countries and territories: an analysis from 1990 to 2025. Sci Rep. 2020;10(1):14790. Published 2020 Sep 8.

Wild

Roglic

Green

, et al. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27(5):1047-1053.

Roper

Bilous

Kelly

, et al.; South Tees Diabetes Mortality Study. Cause-specific mortality in a population with diabetes: South Tees Diabetes Mortality Study. Diabetes Care. 2002;25(1):43-48.

Roglic

Unwin

. Mortality attributable to diabetes: estimates for the year 2010. Diabetes Res Clin Pract. 2010;87(1):15-19.

Grundy

Benjamin

Burke

, et al. Diabetes and cardiovascular disease: a statement for healthcare professionals from the American Heart Association. Circulation. 1999;100(10):1134-1146.

Khalaf

Ibrahim

Bedair

, et al. Plasminogen activator inhibitor-1 gene polymorphism as a risk factor for vascular complications in type 2 diabetes mellitus. Egypt J Med Hum Genet. 2019;20:18.

Jay

Hitomi

Griendling

. Oxidative stress and diabetic cardiovascular complications. Free Radic Biol Med. 2006;40(2):183-192.

Lontchi-Yimagou

Sobngwi

Matsha

, et al. Diabetes mellitus and inflammation. Curr Diab Rep. 2013;13(3):435-444.

10.

Marques-Vidal

Schmid

Bochud

, et al. Adipocytokines, hepatic and inflammatory biomarkers and incidence of type 2 diabetes. The CoLaus study. PLoS One. 2012;7(12):e51768.

11.

Bosevski

Stojanovska

Apostolopoulos

. Inflammatory biomarkers: impact for diabetes and diabetic vascular disease. Acta Biochim Biophys Sin (Shanghai). 2015;47(12):1029-1031.

12.

Strissel

Stancheva

Miyoshi

, et al. Adipocyte death, adipose tissue remodeling, and obesity complications. Diabetes. 2007;56(12):2910-2918.

13.

Vazzana

Ranalli

Cuccurullo

, et al. Diabetes mellitus and thrombosis. Thromb Res. 2012;129(3):371-377.

14.

Petrauskiene

Falk

Waernbaum

, et al. The risk of venous thromboembolism is markedly elevated in patients with diabetes. Diabetologia. 2005;48(5):1017-1021.

15.

Gealekman

Guseva

Hartigan

, et al. Depot-specific differences and insufficient subcutaneous adipose tissue angiogenesis in human obesity [published correction appears in Circulation. 2019 Sep 17;140(12):E693]. Circulation. 2011;123(2):186-194.

16.

Akash

MSH

Rehman

Liaqat

. Tumor necrosis factor-alpha: role in development of insulin resistance and pathogenesis of type 2 diabetes Mellitus. J Cell Biochem. 2018;119(1):105-110.

17.

Ridker

Silvertown

. Inflammation, C-reactive protein, and atherothrombosis. J Periodontol. 2008;79(8 Suppl):1544-1551.

18.

Kristiansen

Mandrup-Poulsen

. Interleukin-6 and diabetes: the good, the bad, or the indifferent? Diabetes. 2005;54(Suppl 2):S114-S124.

19.

King

Mainous 3rd

Buchanan

Pearson

. C-reactive protein and glycemic control in adults with diabetes. Diabetes Care. 2003;26(5):1535-1539.

20.

Samaropoulos

Light

Ambrosius

, et al. The effect of intensive risk factor management in type 2 diabetes on inflammatory biomarkers. Diabetes Res Clin Pract. 2012;95(3):389-398.

21.

Bembde

. A study of plasma fibrinogen level in type-2 diabetes mellitus and its relation to glycemic control. Indian J Hematol Blood Transfus. 2012;28(2):105-108.

22.

Pradhan

Everett

Cook

, et al. Effects of initiating insulin and metformin on glycemic control and inflammatory biomarkers among patients with type 2 diabetes: the LANCET randomized trial. JAMA. 2009;302(11):1186-1194.

23.

Sobel

. Effects of glycemic control and other determinants on vascular disease in type 2 diabetes. Am J Med. 2002;113(Suppl 6A):12S-22S.

24.

Qaseem

Wilt

Kansagara

, et al. Hemoglobin A1c targets for glycemic control with pharmacologic therapy for nonpregnant adults with type 2 diabetes mellitus: a guidance statement update from the American College of Physicians. Ann Intern Med. 2018;168(8):569-576.

25.

Movahed

Hashemzadeh

Jamal

. The prevalence of pulmonary embolism and pulmonary hypertension in patients with type II diabetes mellitus. Chest. 2005;128(5):3568-3571.

26.

Germolec

Shipkowski

Frawley

, et al. Markers of inflammation. Methods Mol Biol. 2018;1803:57-79.

27.

Speidl

Zeiner

Nikfardjam

, et al. An increase of C-reactive protein is associated with enhanced activation of endogenous fibrinolysis at baseline but an impaired endothelial fibrinolytic response after venous occlusion. J Am Coll Cardiol. 2005;45(1):30-34.

28.

Nkansah

Addai-Mensah

Mensah

, et al. Plasminogen activator inhibitor-1 in poorly controlled vs well controlled type-2 diabetes mellitus patients: a case-control study in a district hospital in Ghana. PLoS One. 2021;16(4):e0250090.

29.

Vischer

. von Willebrand factor, endothelial dysfunction, and cardiovascular disease. J Thromb Haemost. 2006;4(6):1186-1193.

30.

Kantarcioglu

Darki

Siddiqui

, et al. The relevance of anti-PF4 antibody isotypes and endogenous glycosaminoglycans and their relationship with inflammatory biomarkers in pulmonary embolism patients. Clin Appl Thromb Hemost. 2022;28:10760296221091770.

31.

Sowers

Lester

. Diabetes and cardiovascular disease. Diabetes Care. 1999;22(Suppl 3):C14-C20.

32.

Ceriello

Giacomello

Stel

, et al. Hyperglycemia-induced thrombin formation in diabetes. The possible role of oxidative stress. Diabetes. 1995;44(8):924-928.

33.

R Charlier

Meier

Jick

Meier

Becker

. Association between glycemic control and risk of venous thromboembolism in diabetic patients: a nested case-control study. Cardiovasc Diabetol. 2022;21(1):2.

34.

Lerstad

Brodin

Enga

, et al. Hyperglycemia, assessed according to HbA1c, and future risk of venous thromboembolism: the Tromsø study. J Thromb Haemost. 2014;12(3):313-319.

35.

Bell

Selvin

Lutsey

, et al. Glycemia (hemoglobin A1c) and incident venous thromboembolism in the atherosclerosis risk in communities cohort study. Vasc Med. 2013;18(5):245-250.

36.

Fangel

Nielsen

Kristensen

, et al. Glycemic status and thromboembolic risk in patients with atrial fibrillation and type 2 diabetes mellitus: A Danish cohort study. Circ Arrhythm Electrophysiol. 2019;12(5):e007030.

37.

Isordia-Salas

Galván-Plata

Leaños-Miranda

, et al. Proinflammatory and prothrombotic state in subjects with different glucose tolerance status before cardiovascular disease. J Diabetes Res. 2014;2014::631902.

38.

Steiner

Reinhardt

Krammer

, et al. Increased levels of soluble adhesion molecules in type 2 (non-insulin dependent) diabetes mellitus are independent of glycaemic control. Thromb Haemost. 1994;72(6):979-984.

39.

Knöbl

Schernthaner

Schnack

, et al. Haemostatic abnormalities persist despite glycaemic improvement by insulin therapy in lean type 2 diabetic patients. Thromb Haemost. 1994;71(6):692-697.

40.

Rosove

Frank

Harwig

. Plasma beta-thromboglobulin, platelet factor 4, fibrinopeptide A, and other hemostatic functions during improved, short-term glycemic control in diabetes mellitus. Diabetes Care. 1984;7(2):174-179.

41.

Osende

Badimon

Fuster

, et al. Blood thrombogenicity in type 2 diabetes mellitus patients is associated with glycemic control. J Am Coll Cardiol. 2001;38(5):1307-1312.

42.

Jiménez-García

Albaladejo-Vicente

Hernandez-Barrera

, et al. Type 2 diabetes is a risk factor for suffering and for in-hospital mortality with pulmonary embolism. A population-based study in Spain (2016-2018). Int J Environ Res Public Health. 2020;17(22):8347.

43.

de Miguel-Díez

Muñoz-Rivas

Jiménez-García

, et al. Type 2 diabetes is associated with a higher incidence of hospitalization for pulmonary embolism in Spain: analysis of hospital discharge data during 2004-2013. Respirology. 2016;21(7):1277-1284.

44.

de Miguel-Díez

López-de-Andrés

Jiménez-Trujillo

, et al. Mortality after pulmonary embolism in patients with diabetes. Findings from the RIETE registry. Eur J Intern Med. 2019;59:46-52.