Circulating Angiogenic Factors and Ischemic Diabetic Foot Syndrome Advancement—A Pilot Study

Despite clear evidence of inadequate angiogenesis in ischemic diabetic foot syndrome (DFS) pathogenesis, angiogenic factor level changes in patients with ischemic DFS remain inconsistent. This study aimed to assess circulating angiogenic factors concerning ischemic DFS advancement and describe their relationships with patients’ clinical characteristics, microvascular parameters, and diabetic control. The study included 41 patients with ischemic DFS (67.3 (8.84) years; 82.9% males). Angiogenic processes were assessed by identifying circulating concentrations of five pro- and two anti-angiogenic factors. We found that penetrating ulcers were related to a significantly higher FGF-2 level (8.86 (5.29) vs. 5.23 (4.17) pg/mL, p = 0.02). Moreover, plasma FGF-2 showed a significant correlation with the SINBAD score (r = 0.32, p = 0.04), platelet count (r = 0.43, p < 0.01), white cell count (r = 0.42, p < 0.01), and age (r = −0.35, p = 0.03). We did not observe any significant linear relationship between the studied biomarkers and microcirculatory parameters, nor for glycemic control. In a univariate analysis using logistic regression, an increase in plasma FGF-2 was tied to greater odds of high-grade ulcers (OR 1.16; 95% CI 1.02–1.38, p = 0.043). This suggests that circulating FGF-2 may serve as a potential biomarker for predicting DFU advancement and progression. It is necessary to conduct further studies with follow-up observations to confirm this hypothesis.


Introduction
Diabetic foot syndrome (DFS) is one of the most common, chronic, and complex complications of diabetes mellitus (DM). It is estimated that the pooled worldwide prevalence of diabetic foot ulcerations (DFUs) is 6% [1] and the lifetime risk of DFUs in diabetes patients even reaches 34% with an annual incidence of 2% [2]. Based on the International Working Group of the Diabetic Foot (IWGDF), a diabetic foot may be characterized by infection, ulceration, or destruction of foot tissue in a person with diagnosed diabetes [3]. The pathophysiology of DFS is complex, but the underlying mechanism can be presented Despite compelling evidence suggesting inadequate angiogenesis and microvascular dysfunction in the pathogenesis of diabetic foot syndrome, the alterations in angiogenic factor levels in the bloodstream or wound materials among individuals with DFUs remain inconsistent [17]. Because of the clinical heterogeneity of ulcerations, accurately monitoring the healing process and predicting the probability of limb amputation poses a challenge. To date, there are no reliable risk prediction models. Therefore, developing such a model or deriving biochemical predictors of poor prognosis is highly interesting.
This study aimed to assess circulating pro-and anti-angiogenic factors concerning DFS advancement and describe potential relationships between clinical characteristics, microcirculatory parameters, and diabetic control among patients with ischemic DFS.

Study Design and Population
A single-centered, cross-sectional study involving patients recruited from the Clinical Department of Angiology between February 2021 and May 2022 was conducted. The inclusion criteria were as follows: (1) subjects aged 40-80 years old with DM type 2, (2) ischemic DFS with active lower limb ulcerations, and (3) concomitant critical limb ischemia due to PAD (category 5 or 6 in the Rutherford classification). Patients with myocardial infarction or stroke within the last 6 months, diagnosed Charcot's foot, acute lower limb ischemia within the previous 3 months, chronic kidney disease with eGFR < 45 mL/min/1.73 m 2 , or neoplasm diagnosed within 5 years were excluded from the study. The study did not include subjects with chronic infectious diseases (e.g., hepatitis C virus, or human immunodeficiency virus) or autoimmune comorbidities (e.g., rheumatoid arthritis, or scleroderma).

Data Collection
Information on chronic disorders, smoking status, and medications was obtained during medical history collection. All data and results were stored using a certificated tool for data collection (Medrio EDC). Hypertension was defined as an SBP ≥ 140 mm Hg and/or DBP ≥ 90 mm Hg, or receiving antihypertensive treatment. Body mass index (BMI) was calculated by dividing weight in kilograms and height in meters squared. Hemoglobin level, platelet count (PLT), white blood cell count (WBC), creatinine, C-reactive protein (CRP), glycated hemoglobin (HbA1c), total cholesterol (TC), triglyceride (TG), highdensity lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C) concentrations were measured on the day of admission to the hospital and were extracted from the hospital laboratory system. The extra blood samples were collected from all patients and tested for the biomarker levels listed below. All samples were collected after a 12 h fasting period.

Assessment of DFS Advancement
The category of chronic limb ischemia was evaluated using the symptomatic Rutherford classification, which defines category 5 as minor tissue loss and category 6 as major tissue loss extending above the transmetatarsal level [18]. The SINBAD classification system and the WIfI classification developed by the Society of Vascular Surgery were used to assess diabetic foot syndrome advancement and amputation risk estimation. WIfI comprises an evaluation of the wound presence, insensitivity of ischemia, and foot infection with 4 grades of severity for each category. It can also determine 64 possible clinical combinations with an estimated risk of amputation at 1 year (very low, low, moderate, or high) [19]. In the SINBAD classification, six elements are graded: ulcer site (forefoot vs. midfoot/hindfoot), ischemia (at least one pulse palpable vs. evidence of ischemia), neuropathy (absent vs. present), bacterial infection (absent vs. present), area (ulcer < 1 cm 2 vs. ≥ 1 cm 2 ), and depth (confined to the skin and subcutaneous tissue vs. reaching muscle, tendons, or more profound). The components of the SINBAD classification can be summed to produce a score between 0 and 6. It divides ulcers into 3 groups: low grade (0-2), moderate grade (3)(4), and high grade (5)(6) [20].

Hemodynamic Parameters of Lower Limb Arteries
The status of lower limb ischemia was assessed using an estimated ankle-brachial index (ABI) and toe-brachial index (TBI). Both examinations included systolic blood pressure measurements of the brachial arteries, ankle arteries, and toe capillaries. These were performed using a sphygmomanometer, an 8 MHz blind Doppler flow detector, and digital plethysmography. ABI was determined by dividing the higher systolic blood pressure measured at the dorsalis pedis or posterior tibial artery and the higher systolic blood pressure measured at the right or left brachial artery. In addition to ABI, systolic blood pressure on the toe was measured with a plethysmographic sensor and TBI was estimated. TBI was assessed by dividing the systolic blood pressure measured at the toe capillaries and the higher systolic blood pressure at the right or left brachial artery. All examinations were performed in a controlled environment with a room temperature between 21 and 23 degrees Celsius. The measurements were taken after a 15 min rest in a supine position with the limbs parallel to the body.

Microcirculation Assessment
Microcirculation was evaluated using laser Doppler flowmetry (LDF) and the transcutaneous oximetry test (tcpO2). Both measurements were made with the Periflux 6000 (Perimed AB, Järfälla, Sweden), equipped with thermostatic laser Doppler probes to precisely heat the tissue at the measurement site and a modified Clark's polarographic oxygen sensor. The probes were stabilized on the dorsal aspect of the distal part of the foot, excluding skin with necrosis or inflammation, bone prominence, or a superficial tendon. Baseline microvascular blood flow was measured for 5 min and expressed in arbitrary perfusion units (PU). The tcpO2 examination was conducted for 20 min or until the curve on the graph flattened and was presented in mmHg. The results reflect microvascular flow, including capillaries, arterioles, venules, and shunts, as well as microcirculation blood perfusion, metabolic activity, oxyhemoglobin dissociation, and tissue oxygen partial pressure. Both tests were conducted in a room at a temperature of 21-23 degrees Celsius. Testing was preceded by a 15 min resting period, during which the patient lay in a comfortable position.

Statistical Analysis and Sample Size Calculation
Statistical analysis was performed in R 4.2.2 (R Team, R Statistical Foundation, Vienna, Austria) using publicly available packages (tidyverse, arsenal, rstatix, and ggpubr). Continuous variables were summarized as mean and standard deviation, with categorical variables as counts and proportions. Right-skewed variables were log-transformed before analyses. Distribution was assessed using qqplots. Continuous and nominal variables were compared using a t-test and Fisher's test, respectively. Pearson's correlation coefficient was used to assess linear relationships. Tests were two-tailed and a p-value < 0.05 was treated as significant.
We assumed a moderate (r = 0.2 to r = 0.4) association between circulating angiogenesis regulating factors and clinical characteristics. Given a significance level of 0.05 and a power set at 0.80, the required sample size for a one-sided hypothesis is at least 37 patients.

Ethical Aspects
This research is not in conflict with any ethical norms and regulations in research studies on humans. All patients signed informed consent forms to participate in the study. The regulations of the Declaration of Helsinki were used to prepare the protocol for this study, and the proper consent for this study was obtained from the constituted committee for human subjects or animal research at Jagiellonian University Medical College (decision number 1072.6120.129.2020).

Comparison of Subjects' Characteristics According to Ulcer Features
The comprehensive clinical characteristics of subjects divided into two groups based on ulceration depth (limited vs. penetrating) are presented in Table S1. A comparison of blood count and biochemical tests between the groups showed that patients with penetrating ulcers had a significantly lower hemoglobin level (13.79 (2.10) vs. 12.28 (1.66), p = 0.01).

Comparison of Angiogenic Factors and Clinical Characteristics According to Patients' Age
Generally, age did not significantly impact most of the clinical features (Table S2). However, in patients aged 40-68, the levels of FGF-2 were significantly higher compared with those in patients over 68 years old (

Relationships between Pro-and Anti-Angiogenic Factors, Microvascular Status, and Glycemic Control
We did not observe any significant linear relationship between the studied angiogenic biomarkers and microcirculatory parameters (LDF, tcpO2) (Figures 2, 4, 6, 8, 10, 12, and 14). There was also no impact from glycemic control (defined as HbA1c ≤ 8% or HbA1c > 8%) on the concentrations of the angiogenic markers, as shown in Figure 15.

Relationships between Plasma FGF-2 and Clinical Characteristics
We found a significant correlation between circulating FGF-2 and several clinical characteristics (Figures 1 and 2). It was associated with penetrating ulcer presence (mean (SD); 8.86 (5.29)                             In the univariate analysis using logistic regression, an increase in plasma FGF-2 was tied to greater odds of high-grade ulcers according to the SINBAD classification (OR 1.16; 95% CI 1.02-1.38, p = 0.043). Using a multivariable model with log-transformed CRP and FGF-2 concentrations, we observed an improvement in model fit. However, only the relationship with CRP remained significant (OR 1.98; 95% CI 1.15-3.91, p = 0.025). We further assessed the predictive performance of this simple model using three thresholds (0.3, 0.5, and 0.7), which yielded a corresponding sensitivity and specificity of 94.74%/27.78%, 68.42%/72.22%, and 42.11%/88.89%, respectively.

Discussion
The key finding of the present study is that FGF-2 was significantly associated with ulceration depth and showed a significant correlation with the SINBAD score. We found that the mean FGF-2 level was higher among subjects with penetrating ulcers and a higher SINBAD score. Moreover, an increase in plasma FGF-2 was tied to greater odds of highgrade ulcers. In this study, among the analyzed angiogenic factors, only plasma FGF-2 In the univariate analysis using logistic regression, an increase in plasma FGF-2 was tied to greater odds of high-grade ulcers according to the SINBAD classification (OR 1.16; 95% CI 1.02-1.38, p = 0.043). Using a multivariable model with log-transformed CRP and FGF-2 concentrations, we observed an improvement in model fit. However, only the relationship with CRP remained significant (OR 1.98; 95% CI 1.15-3.91, p = 0.025). We further assessed the predictive performance of this simple model using three thresholds (0.3, 0.5, and 0.7), which yielded a corresponding sensitivity and specificity of 94.74%/27.78%, 68.42%/72.22%, and 42.11%/88.89%, respectively.

Discussion
The key finding of the present study is that FGF-2 was significantly associated with ulceration depth and showed a significant correlation with the SINBAD score. We found that the mean FGF-2 level was higher among subjects with penetrating ulcers and a higher SINBAD score. Moreover, an increase in plasma FGF-2 was tied to greater odds of highgrade ulcers. In this study, among the analyzed angiogenic factors, only plasma FGF-2 concentrations could differentiate patients with penetrating ulceration. These results may suggest a hypothesis that elevated systemic FGF-2 could promote more advanced DFU. To the best of our knowledge, currently, there is a lack of research regarding the significance of FGF-2 as a predictor in the advancement and progression of DFS.
Our results are consistent with existing data on circulating FGF-2 concentration differences between DFU patients and healthy subjects. Kulwas et al. showed increased circulating FGF-2 levels in DFU compared with individuals without DFU [21]. Moreover, in Tecilazih's study, median values of serum FGF-2 were the highest for DFU patients, but no statistical analysis was conducted, and, thus, no definite conclusions can be drawn [22]. Gui et al. obtained comparable findings, as they noted an increase in the concentration of circulating FGF-2 among diabetic individuals in contrast with the control group [23]. However, these findings are in contrast with the general knowledge about FGF-2 angiogenic and mitogenic effects, as well as its role in dermal fibroblast migration and stimulating the proliferation phase in wound healing [24]. FGF-2 is one of the most widely studied factors in patients with DFS, and its use as a topically administered drug in non-healing ulcers has been the main topic in recent trials [24,25].
On the other hand, circulating FGF-2 plays a role in inflammatory processes and atherosclerotic lesion growth. It may stimulate intimal thickening, intraplaque angiogenesis, and the proliferation of vascular smooth muscle cells [26,27]. Furthermore, local hyperglycemia and advanced glycation end products (AGEs) in diabetic patients may promote glycosylated FGF-2 production, which inhibits the proliferation of endothelial cells and has negative effects on wound healing [21,28]. The glycation of angiogenic factors is associated with unresponsiveness to them [29]. One study showed that increased plasma FGF-2 levels might even be a risk biomarker for coronary heart disease occurrence in adult men with DM type 2 [30]. Therefore, it is crucial to differentiate between its various effects depending on the analyzed material. While its local deficiency in healing tissue is associated with insufficient angiogenesis and delayed healing, its elevated circulating form may have unfavorable effects on the course of DFS and cardiovascular disease progression in patients with DM. The potential double pathway of FGF-2 action on impaired wound healing is presented in Figure 16.
On the other hand, circulating FGF-2 plays a role in inflammatory processes and atherosclerotic lesion growth. It may stimulate intimal thickening, intraplaque angiogenesis, and the proliferation of vascular smooth muscle cells [26,27]. Furthermore, local hyperglycemia and advanced glycation end products (AGEs) in diabetic patients may promote glycosylated FGF-2 production, which inhibits the proliferation of endothelial cells and has negative effects on wound healing [21,28]. The glycation of angiogenic factors is associated with unresponsiveness to them [29]. One study showed that increased plasma FGF-2 levels might even be a risk biomarker for coronary heart disease occurrence in adult men with DM type 2 [30]. Therefore, it is crucial to differentiate between its various effects depending on the analyzed material. While its local deficiency in healing tissue is associated with insufficient angiogenesis and delayed healing, its elevated circulating form may have unfavorable effects on the course of DFS and cardiovascular disease progression in patients with DM. The potential double pathway of FGF-2 action on impaired wound healing is presented in Figure 16.  We observed that FGF-2 concentrations were more likely to be higher among patients who were ever smokers (active and past smokers). These results are consistent with observations obtained from bovine models. In aortic endothelial and smooth muscle cells, bFGF release after nicotine stimulation was greater than in controls [31,32]. Similar results were obtained from previous studies using human osteoblasts and smooth muscle cells [33,34]. We showed that plasma FGF-2 significantly correlated with PLT and WBC. These findings align with a study on platelet activation in burn patients, where the amount of FGF-2 per platelet was constant for 21 days [35]. The FGF family has a known influence on inflammatory processes [36]. A few studies have shown that FGF-2 is upregulated in inflammatory disorders and may act as an immunoregulator of T-lymphocytes, neutrophils, macrophages, and monocytes [36][37][38]. These findings may support our result regarding the increase in FGF-2 levels with WBC.
Additionally, we presented a reverse relationship between FGF-2 and age. This phenomenon remains a subject of discussion according to Harely M. et al., who found a progressive decrease in FGF-2 with age in human mesenchyme-derived progenitor cell cultures [39]. However, there are also studies in which relationships between FGF-2 levels and age were not present [21,40].
Although there is convincing evidence indicating insufficient angiogenesis in the development of DFS, the changes in circulating angiogenic factor levels in the blood or wound materials among patients with DFUs are inconclusive. Several studies have emphasized the significance of reduced levels of pro-angiogenic factors such as VEGF-A and FGF-2 in wound material, as well as the decreased expression of PlGF in the development of DFUs [17,41]. Additionally, research has validated that increased levels of anti-angiogenic factors, such as PEDF, play a significant role in impeding angiogenesis and the healing of wounds [42]. Numerous inconsistencies in the field remain, such as the paradox of impaired wound healing despite the presence of high levels of pro-angiogenic factors or fluctuations in the expression of pro-angiogenic factors during the healing process.
According to ulceration features, our study found that ulceration depth had significant relationships with certain clinical characteristics. Penetrating ulcers were tied to lower hemoglobin levels, higher platelet counts, and CRP concentrations, which are reflective of an inflammatory state with chronic iron restriction. Our observation is consistent with conclusions obtained from the other studies. A recent systematic review showed that lower hemoglobin levels are associated with DFU advancement, non-healing ulcers, amputation, and mortality [43]. Further, Wang et al. showed that CRP is an essential biomarker in differentiating grade 1 from grade 2 DFUs [44]. CRP has been proposed as a prognostic marker for DFU healing [22].
Only a few studies have described the effect of HbA1c levels on ulcer healing [45,46]. Chen et al. showed that HbA1C might be one of the risk factors for early DFU development in diabetic patients [47]. Nevertheless, we did not observe any significant correlations between the studied biomarker panel and glycemic control, nor microcirculatory parameters. While a potential linear trend between vascular endothelial growth factor receptor 2 (VEGF-R2) and both tcpO2 and LDF measurements was observed, whether such a relationship exists requires confirmation in larger samples.
The main strength of this study is a relatively large number of assessed circulating angiogenic factors, which were evaluated in a homogeneous group of patients with ischemic DFS. Despite concerns regarding study power, we observed that FGF-2 might be tied to specific clinical characteristics reflecting DFS advancement. Nevertheless, our study does have several limitations. The main one includes the small sample size, the monocentric character of the study, and the lack of a healthy control group. This cross-sectional study examined forty-one ischemic DFS patients who represent a homogenous sample because of highly restrictive recruitment criteria (in contrast to earlier studies [20,21]). Nevertheless, as a pilot study, it enabled the initial identification of the most promising angiogenic factor in terms of its predictive potential.

Conclusions
This study indicated a significant correlation between elevated plasma FGF-2 levels and ulceration depth, as well as with the SINBAD score, in ischemic DFS patients. Penetrating ulcers were related to significantly higher plasma FGF-2 concentrations, and an increase in FGF-2 was tied to greater odds of high-grade ulcerations. This suggests that FGF-2 may serve as a potential biomarker for predicting DFU advancement and progression. Future research with follow-up observations should investigate changes in circulating FGF-2 over time and their relationships with DFU healing and lower limb amputation rates to verify its predictive value.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biomedicines11061559/s1, Table S1. Comparison of clinical characteristics according to ulcer depth. Data are presented as mean (SD) unless otherwise indicated; Table S2: Comparison of angiogenic factors and clinical characteristics according to patient ages.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Ethics Committee of Jagiellonian University (protocol code 1072.6120.129.2020).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: Data sharing is not applicable to this article.