Early Dysregulation of Angiopoietin-1 and -2 as a Predictor of Mortality in Critically Ill Burn Patients

Abstract

Shock-induced endothelial dysfunction plays a critical role in burn pathophysiology, with endothelial glycocalyx layer degradation promoting systemic inflammation, vascular instability, and multi-organ failure. The angiopoietin-tunica interna endothelial cell kinase (TIE2) axis, particularly the angiopoietin-1 (Ang-1) and angiopoietin-2 (Ang-2) balance, regulates endothelial function; elevated Ang-2 and a high Ang-2/1 ratio are linked to worse outcomes in critical illness. While well-documented in sepsis and trauma, effects of burn-induced angiopoietin dysregulation remain unclear. This study evaluates Ang-1, Ang-2, and the Ang-2/1 ratio as biomarkers of endothelial dysfunction and predictors of 30-day mortality in patients with burn injuries. In this prospective study, 62 adult patients with burn injuries were enrolled (January 2021–November 2024), with serum Ang-1 and Ang-2 measured via enzyme-linked immunosorbent assay on postburn day 1. Of 62 patients, 52 were analyzed; 78.05% of survivors and 90.91% of non-survivors were male. Median age was 45 (survivors) vs 54 years (non-survivors, P = .139). Non-survivors trended toward burns > 20% TBSA (72.73% vs 41.46%, P = .093). Ang-1 was lower in non-survivors (3.96 vs 7.97 ng/mL, P < .001), predicting early mortality (area under the receiver operating characteristic [AUROC]: 0.82) with a cut-off of 4.825 ng/mL and decreased mortality risk (odds ratio [OR]: 0.63, 95% confidence interval [CI]: 0.40-0.87, P = .017). Ang-2 was higher (6.07 vs 1.99 ng/mL, P < .001; AUROC: 0.95), with a cut-off of 3.554 ng/mL. The Ang-2/1 ratio was elevated (1.59 vs 0.23, P < .001; AUROC: 0.93), with a cut-off of 0.504 and increased mortality risk (OR: 2.17, 95% CI: 1.10-5.12, P = .038). Early Ang-1, Ang-2, and Ang-2/1 ratio levels correlate with 30-day mortality and may guide early prognostication.

INTRODUCTION

Patients with severe burns (>20% of total body surface area [%TBSA]) undergo a profound and sustained immune response with extensive clinical sequelae, including multiple surgeries, prolonged intensive care needs, and hard-to-prevent infectious complications (pneumonia, wound infection, and sepsis), culminating in high mortality. The massive systemic inflammation that occurs after burn injury is regularly in interplay with endothelial dysfunction, organ dysfunction, and immune derangements.¹ The impact of shock on endothelial structure and function is a growing area of research in surgical and critical care, particularly in the management of severe burns, trauma, and sepsis. Endotheliopathy, or systemic vascular endothelial dysfunction, is strongly linked to poor outcomes in critically ill patients. This complex process is driven by tissue damage, systemic inflammation, ischemia, and shock, leading to degradation of the endothelial glycocalyx layer (EGL)—a key network of proteoglycans and glycoproteins that preserves vascular integrity, modulates coagulation, and regulates inflammatory responses.^2,3 Shock-induced endotheliopathy (SHINE) is well characterized in hemorrhagic, septic, and cardiogenic shock, where progressive endothelial dysfunction exacerbates vascular permeability, inflammation, and multi-organ failure.^4–6 A hallmark of endotheliopathy is EGL shedding, which contributes to microvascular dysfunction and systemic inflammation.^7,8

Severe burns present a unique challenge to endothelial function. The systemic response to burns triggers profound inflammation, fluid shifts, and compromised skin integrity, increasing the risk of shock, infection, and capillary leak syndrome—a state of widespread endothelial dysfunction, intravascular fluid loss, and tissue edema. Endothelial glycocalyx layer breakdown in patients with burn injuries leads to the systemic release of its key components such as syndecan-1 and tissue factor pathway inhibitor, both associated with increased mortality.^9,10

Vascular endothelial cells maintain the integrity of the EGL. The EGL allows for the free flow of luminal contents in physiological states. The free flow of cells and nutrients allows for the delivery of critical oxygen to perfused organs. Angiopoietin-1 (Ang-1) promotes endothelial stability and quiescence by activating the tunica interna endothelial cell kinase (TIE2) receptor, which strengthens intercellular junctions and reduces vascular permeability. In contrast, angiopoietin-2 (Ang-2) functions as a competitive antagonist to Ang-1, inhibiting TIE2 signaling and destabilizing the endothelium, leading to increased vascular permeability and inflammation.^4,11,12 A dysregulation between Ang-1 and Ang-2, with elevated Ang-2 relative to Ang-1, has been observed in stress states such as postmyocardial infarction¹³ and is associated with poor clinical outcomes including sepsis,¹⁴ acute kidney injury,¹⁵ acute respiratory distress syndrome,^16,17 and mortality.^18,19 In patients meeting systemic inflammatory response syndrome criteria, elevated Ang-2 levels and a high Ang-2/1 ratio are independently associated with mortality, shock, and organ failure.²⁰ Similarly, in trauma patients, these dysregulations correlate with blunt injuries, hemorrhagic shock, and coagulopathy, leading to prolonged ICU stays and increased mechanical ventilation duration.²¹

Despite its established role in endothelial dysfunction across critical illnesses, the contribution of the angiopoietin-TIE2 axis to burn pathophysiology remains poorly defined. A recent study shows a significant rise in the Ang-2/1 ratio within 48 h of severe burns, which correlates with %TBSA, burn severity scores, and in-hospital mortality, suggesting the potential of the Ang-2/1 ratio as a prognostic biomarker in patients with burn injuries, similar to its role in sepsis and trauma.²² Furthermore, the Ang-2/1 ratio correlates with systemic inflammation markers such as C-reactive protein and procalcitonin.^23,24 These findings support the concept that endothelial dysfunction, reflected by the Ang-2/1 ratio, plays a major role in the pathophysiology of severe burns and other critical illnesses. However, prior studies have not clarified whether this dysfunction is primarily driven by elevated Ang-2, suppressed Ang-1, or a combination of both. Furthermore, detailed risk and survival analyses linking these biomarkers to mortality in patients with burn injuries are still lacking.

Building on these findings, this study aims to evaluate Ang-1, Ang-2, and the Ang-2/1 ratio as biomarkers of burn-induced endotheliopathy and predictors of 30-day in-hospital mortality. Understanding the individual and combined contributions of Ang-1 and Ang-2 to endothelial dysfunction is essential for the development of targeted therapies. Identifying angiopoietins as prognostic factors may enhance triage and clinical decision-making. Furthermore, targeting the Ang-TIE signaling axis could represent a novel therapeutic strategy to improve outcomes following severe burn injury.

MATERIALS AND METHODS

Study population

This study was approved by the institutional review board (IRB #215735). All patients aged 18 and older who presented to our institution’s burn center after thermal burn injury were screened for enrollment. Informed consent was obtained from all participants or their legally authorized representatives. All adult patients who sustained burn injuries greater than or equal to 5% TBSA were prospectively enrolled. Patients were excluded if they had prior cancer, hematologic or immunologic diseases, concurrent injury, or if mortality was expected within 24 h of admission based on the clinical judgment of the treating attending physician. Between January 2021 and November 2024, a total of 62 patients met the criteria and provided consent for enrollment. A total of 52 patients were included in the overall analysis due to missing postburn day 1 (PBD1) blood draws. For the Ang-2 and Ang-2/Ang-1 ratio analysis, 3 patients were excluded due to failed quantification of Ang-2 during enzyme-linked immunosorbent assay (ELISA) processing, resulting in 49 patients (40 survivors, 9 non-survivors) included in this subanalysis. Ang-1 was successfully quantified in all samples, with values falling within established quality control parameters (Figure 1). The most common reason for exclusion was updated time from injury after further scene information was obtained. These exclusions were made if the PBD1 blood sample was collected < 18 h or > 30 h after confirmation of burn injury. Among the included patients, there was no significant difference in average time to blood draw between survivors and non-survivors.

Figure 1.

Study Population Participation. From January 2021 to November 2024, 62 Patients With Burn Injuries (≥5% TBSA) Consented to Enrollment. Ang, angiopoietin. Created in BioRender. Kubasiak, J. (2025) https://BioRender.com/fj01vfr.

Clinical patient data

Patient demographics, laboratory results, and treatment data were prospectively collected from medical records for all patients included in the study. Demographic data included age, gender, race, and body mass index (BMI). Clinical data included %TBSA, length of hospital and ICU stay, need for mechanical ventilation, and the number of ventilator days for patients requiring mechanical ventilation, based on definitions from the Burn Care Quality Platform (BCQP) Registry data dictionary.²⁵ Inhalation injury (IHI) was diagnosed based on bronchoscopy findings, with criteria including the presence of soot, airway edema, or mucosal injury.

Sample collection

Peripheral blood samples were collected on PBD1 (defined as 18-30 h postburn) in SST Serum Separation Tubes (BD Biosciences, Franklin Lakes, NJ) after the patient had undergone approximately 24 h of institutional protocol resuscitation (2 mL × body weight in kg × %TBSA burned). Samples were centrifuged at 2500g for 10 min according to manufacturer recommendations to isolate serum. Aliquots were generated and frozen at −80 °C until batch analysis.

Biomarker measurements

Serum Ang-1 and Ang-2 levels were quantified using precoated ELISA kits (Ang-1: EHANGPT1; Ang-2: KHC1641; Thermo Fisher Scientific, Waltham, MA), following the manufacturer’s protocols. Samples were diluted 1:10 and applied to wells coated with capture antibodies specific to Ang-1 or Ang-2. Standard curves were generated using serial dilutions of recombinant proteins. Optical density was measured for all samples and standards in technical duplicates using a SpectraMax iD5 microplate reader (Molecular Devices, San Jose, CA).

Statistical analysis

We correlated Ang-1, Ang-2, and Ang-2/1 levels with the primary outcome measure in-hospital mortality within 30 days of admission. In addition, we correlated Ang-1, Ang-2, and Ang-2/1 levels with the secondary outcomes burn severity (%TBSA) and requirement for mechanical ventilation. Follow-up time was defined as 30 days from hospital admission, with patient outcomes classified as death or survival within the 30-day period. Burn severity was stratified as low if %TBSA was ≤ 20% and high if %TBSA was > 20%. The predictive ability of Ang-1, Ang-2, and the Ang-2/1 ratio measured on PBD1 for 30-day mortality was evaluated using receiver operating characteristic (ROC) curves. For comparison, an ROC curve was also generated for TBSA. Optimal cut-off values for PBD1 Ang-1, Ang-2, and the Ang-2/1 ratio were determined using the maximum product of sensitivity and specificity, and were set at 4.825 ng/mL, 3.554 ng/mL, and 0.504, respectively.

These cut-off values were established during the study and used to stratify patients into higher or lower biomarker groups for further analysis. Descriptive statistics were used to summarize patient demographics, injury severity, and biomarker levels by 30-day survival status. Categorical variables were presented as frequencies and percentages and compared using the chi-square or Fisher’s exact test (2-sided) as appropriate. Continuous variables were expressed as medians and interquartile ranges (IQRs) and compared between survivors and non-survivors using Mann–Whitney tests (2-sided). Kaplan–Meier survival curves were constructed to assess differences in 30-day survival between groups defined by the Ang-1, Ang-2, and Ang-2/ 1 cut-off values, with significance determined by the log-rank test.

The association between PBD1 Ang-1, Ang-2, and Ang-2/1 levels and 30-day in-hospital mortality was evaluated using logistic regression models to calculate odds ratios (ORs) for the likelihood of mortality with a 95% confidence interval (95% CI). This model was adjusted for age and burn %TBSA. All statistical analyses were performed using R (version 4.3.1, R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism version 10 (GraphPad Software, San Diego, CA). Statistical significance was defined as P < .05.

RESULTS

Demographics

Demographics and injury characteristics are summarized in Table 1. Most patients were male, with 78.05% in the surviving cohort and 90.91% in the non-surviving cohort. The median age was 45 years (IQR: 32-60 years) for survivors and 54 years (IQR: 48-63.5 years) for non-survivors (P = .139). Race distribution was similar across groups, with Caucasians comprising the majority in both survivors (45%) and non-survivors (54.55%). Overall mortality was 21.2% (n = 11). Comorbidities were generally uncommon in the cohort. No patients had chronic kidney disease or coronary artery disease. Diabetes mellitus was present in 17.07% of survivors and 9.09% of non-survivors, while chronic obstructive pulmonary disorder was observed in 7.32% and 18.18%, respectively; neither differed significantly between groups. Burns with TBSA ≤ 20% were considered smaller burns, with TBSA > 20% considered larger burns. Although not statistically significant, non-survivors had a higher proportion of larger burns (72.73%) compared to survivors (41.46%), while 58.54% of survivors and 27.27% of non-survivors had smaller burns (P = .093). Non-survivors had a higher incidence of inhalation injury (81.82% in the non-surviving group vs 9.76% in the surviving group, P < .001). Lengths of stay in both the ICU and overall hospital stay were significantly shorter for non-survivors (P < .001), and non-survivors also had significantly fewer ventilator days (P = .008), which would be expected in those experiencing early mortality due to limited duration of care. All non-survivors required the ventilator, which was a significantly greater proportion than those who survived (P < .0001).

Table 1.

Demographics and Injury Characteristics

	Mortality
Characteristic	Alive, n = 41a	Dead, n = 11a	P valueb
Age, years			.139
Median (IQR)	45 (32, 60)	54 (48, 63.5)
Gender			.668
Male	32 (78.05%)	10 (90.91%)
Female	9 (21.95%)	1 (9.09%)
Race			.777
Asian	3 (7.50%)	0 (0.00%)
Black	8 (20.00%)	1 (9.09%)
Caucasian	18 (45.00%)	6 (54.55%)
Hispanic	9 (22.50%)	4 (36.36%)
Other race	2 (5.00%)	0 (0.00%)
Unknown	1	0
BMI			.552
Median (IQR)	27.37 (23.44, 32.90)	26.71 (25.03, 27.40)
COPD	3 (7.31%)	2 (18.18%)	.283
Diabetes mellitus	7 (17.07%)	1 (9.09%)	>.9999
%Total body surface area (TBSA)			.093
TBSA ≤ 20%, no. (%)	24 (58.54%)	3 (27.27%)
TBSA > 20%, no. (%)	17 (41.46%)	8 (72.73%)
Inhalation injury (IHI)	4 (9.76%)	9 (81.82%)	<.001
LOS, days			<.001
Median (IQR)	20.00 (12.00, 45.00)	4.00 (2.50, 7.50)
ICU	41 (100.00%)	11 (100.00%)
ICU LOS, days			<.001
Median (IQR)	19.00 (11.00, 45.00)	4.00 (3.00, 7.50)
Required ventilator			<.0001
Yes	14 (34.15%)	11 (100.00%)
No	27 (65.85%)	0 (0%)
Ventilator days			.008
Median (IQR)	0.00 (0.00, 5.00)	4.00 (3.00, 7.50)
Day 1 Ang-1 (ng/mL)			<.0001
Median (IQR)	7.97 (5.60, 11.56)	3.96 (1.86, 5.73)
Day 1 Ang-2 (ng/mL)			<.0001
Median (IQR)	1.99 (1.55, 2.47)	6.07 (4.62, 6.39)
Day 1 Ang-2/1			<.0001
Median (IQR)	0.23 (0.14, 0.43)	1.59 (0.60, 3.33)

^a n (%), ^bWilcoxon rank-sum test; Fisher’s exact test; Wilcoxon rank-sum exact test.

Abbreviations: Ang, angiopoietin; BMI, body mass index; COPD, chronic obstructive pulmonary disorder; ICU, intensive care unit; IHI, inhalation injury; LOS, length of stay; PBD, postburn day; POI, point of injury.

Angiopoietin as an endothelial biomarker predicting early mortality

On PBD1, Ang-1 levels were significantly lower in non-survivors compared to survivors (P < .001), with median levels of 3.96 ng/mL (IQR: 1.86, 5.73) in non-survivors and 7.97 ng/mL (IQR: 5.6, 11.56) in survivors. Ang-2 levels were significantly higher in non-survivors (6.07 ng/mL, IQR: 4.62, 6.39) than survivors (1.99 ng/mL, IQR: 1.55, 2.37, P < .001). The Ang-2/1 ratio was also significantly higher in non-survivors (1.59, IQR: 0.60, 3.33) compared to survivors (0.23, IQR: 0.14, 0.43, P < .001). To determine optimal cut-off values for predicting 30-day in-hospital mortality, ROC analysis was performed with subsequent utilization of maximum product of sensitivity and specificity. For Ang-1, the area under the curve was 0.82 (95% CI: 0.6871-0.9581, P = .0011), with an optimal cut-off value at 4.825 ng/mL yielding a sensitivity of 64%, specificity of 78%, positive predictive value (PPV) of 44%, and negative predictive value (NPV) of 89%. For Ang-2, the area under the curve was 0.95 (95% CI: 0.8943-1.000, P < .0001), with an optimal cut-off value at 3.554 ng/mL yielding a sensitivity of 100%, specificity of 90%, PPV of 69%, and NPV of 100%. For the Ang-2/1 ratio, the area under the curve was 0.93 (95% CI: 0.8529-1.000, P < .0001), with an optimal cut-off at 0.504 yielding a sensitivity of 88%, specificity of 83%, PPV of 53%, and NPV of 97%. TBSA was used as a comparison, with an area under the curve resulting in 0.80 (0.6658-0.9403, P = .0017) (Figure 2). Using the above cut-off values, we performed Fisher’s exact test comparing patients above and below the thresholds. Patients with Ang-2 levels and Ang-2/1 ratios above optimal cut-off had significantly increased mortality, while those with Ang-1 levels below optimal cut-off were associated with a lower risk of mortality (Figure 3B, D, and F).

Figure 2.

Receiver Operating Characteristic (ROC) Curves for Angiopoietin-1, -2, and -2/1 Ratio in Comparison With %TBSA. AUC values were 0.82 for Ang-1 (cut-off: 4.825 ng/mL), 0.95 for Ang-2 (cut-off: 3.554 ng/mL), and 0.93 for the Ang-2/1 ratio (cut-off: 0.504), and 0.80 for TBSA. Ang, angiopoietin; AUROC, area under the receiver operating characteristic.

Figure 3.

Postburn Day 1 (PBD1) Biomarker Survival and Contingency Table Analysis. (A, C, E) Patients with Ang-1 ≤ 4.825 ng/mL had lower survival; patients with Ang-2 > 3.554 ng/mL and Ang-2/1 ratio > 0.504 had lower survival. Solid lines signify values below respective optimized cut-off values, dotted lines signify above respective cut-off values. (B, D, F) Higher Ang-1 was associated with reduced mortality; higher Ang-2 and Ang-2/1 ratio was associated with increased mortality.

Likelihood of mortality

Likelihood of mortality was assessed in relation to PBD1 biomarker levels, with adjustments for age and %TBSA (Table 2). PBD1 Ang-1 levels were associated with a reduced likelihood of mortality, with an OR of 0.63 (95% CI: 0.40-0.87; P = .017), suggesting that higher Ang-1 levels on PBD1 decreased the likelihood of mortality. Although not significant, PBD1 Ang-2 did show a trend toward increasing mortality (OR: 1.27, 95% CI: 0.97-1.76; P = .09). Moreover, PBD1 Ang-2/1 ratio was associated with a higher likelihood of mortality (OR: 2.17, 95% CI: 1.10-5.12; P = .038). Kaplan–Meier survival analysis demonstrated that patients with Ang-1 ≤ 4.825 ng/mL had significantly lower survival compared to those with higher levels (P = .026) (Figure 3A). In addition, Ang-2 > 3.554 ng/mL and Ang-2/1 > 0.504 were associated with increased mortality, with highly significant differences observed (P < .0001 for both) (Figure 3C and E). Survival differences were assessed using the log-rank test.

Table 2.

Univariate and Adjusted Logistic Regression Models for 30-Day Mortality

		Unadjusted		Adjusteda
Biomarker		OR (95% CI)	P value	OR (95% CI)	P value
PBD1
	Ang-1	0.63 (0.44, 0.84)	.0048	0.63 (0.40, 0.87)	.017
	Ang-2	1.40 (1.12, 2.07)	.03	1.27 (0.97, 1.76)	.09
	Ang-2/1	2.598 (1.44-6.85)	.0121	2.17 (1.10, 5.12)	.038

^aAdjusted for age and %TBSA.

Abbreviations: Ang, angiopoietin; CI, confidence interval; OR, odds ratio.

Association between secondary outcomes and biomarkers of endothelial dysfunction and injury severity

Non-survivors trended toward a higher proportion of larger burns (72.73%) compared to survivors (41.46%), while survivors seemed to more often have smaller burns (58.54%) compared to non-survivors (27.27%) (P = .093; Table 1). For Ang-1, a significant association was found, such that higher TBSA patients were more likely to have Ang-1 levels ≤ 4.825 ng/mL (P = .04; Figure 4A). For Ang-2, a significant relationship was also observed, such that higher TBSA patients were more likely to have higher Ang-2 levels > 3.554 ng/mL (P < .0001; Figure 4C). Similarly, the Ang-2/1 ratio showed a significant association, such that higher TBSA patients had higher Ang-2/1 ratios > 0.504 (P = .0012; Figure 4E). Patients with higher Ang-1 levels were less likely to require a ventilator (P = .0002), while those with higher Ang-2 levels and higher Ang-2/1 ratios were more likely to need a ventilator (P < .0001) (Figure 4B, D, and F).

Figure 4.

Postburn Day 1 (PBD1) Biomarker Secondary Outcome Contingency Table Analysis. PBD1 Ang-1, Ang-2, and Ang-2/1 ratio association with burn size and mechanical ventilation. Higher and lower Ang-1, Ang-2, and Ang-2/1 are defined by cut-off values of 4.825 ng/mL, 3.554 ng/mL, and 0.504, respectively. Patients with higher TBSA (>20%) were more likely to have low Ang-1 (P = .04) (A), high Ang-2 (P < .0001) (C), and high Ang-2/1 ratio (P = .0012) (E). Higher Ang-1 was associated with reduced need for ventilation (P = .0002) (B), while higher Ang-2 and Ang-2/1 were associated with increased need (P < .0001) (D, F). Ang-1, angiopoietin-1; Ang-2, angiopoietin-2.

DISCUSSION

Primary outcome

This study evaluates the role of Ang-1, Ang-2, and the Ang-2/1 ratio in predicting early mortality in critically ill patients with burn injuries. By analyzing these biomarkers 24 h after the burn injury, we capture the postresuscitation phase, minimizing any factors from under resuscitation in the field. Our findings show that Ang-1 is significantly higher in survivors compared to non-survivors. After adjusting for possible confounders including age and burn size, increasing Ang-1 levels are associated with a decreased likelihood of early mortality. In addition, survival analysis demonstrated that patients with Ang-1 ≤ 4.825 ng/mL had significantly lower survival. These findings support the hypothesis of its protective role in critically ill patients with burn injuries. Conversely, Ang-2 levels were significantly elevated in non-survivors. Although adjusted logistic regression analysis did not result in a significant finding, Ang-2 demonstrated a trend toward increased mortality. Furthermore, survival analysis confirmed that Ang-2 levels above the threshold were strongly associated with lower survival. Similarly, our Ang-2/1 ratio aligns with current literature²² in showing significantly higher levels in non-survivors, with elevated ratios correlating with an increased risk of mortality after adjusting for age and burn severity. Survival analysis also resulted in a significant correlation with mortality above the optimized cut-off value (>0.504). In summary, this pilot study revealed that decreased Ang-1 levels, increased Ang-2 levels, and a higher Ang-2/1 ratio on PBD1 are associated with increased mortality after severe burn injury, with the strongest correlation found through the Ang-2/1 ratio (OR: 2.17, P = .038). As these levels are readily assessed on routine blood draws, consideration for incorporation into post resuscitation prognostication is recommended.

Literature commentary

Currently, we are aware of only one other publication that presents findings on the prognostic ability of angiopoietins following severe burns. In a similarly sized cohort (n = 56), Heuberger et al. characterized the Ang-2/1 ratio for ICU survivors and non-survivors. They demonstrated an increase in the Ang-2/1 ratio over time, correlating with increased TBSA, burn severity, and in-hospital mortality.²² These results align with ours; however, beyond presenting the Ang-2/1 ratio distribution among survivors and non-survivors, we also provide detailed findings for each protein level, along with a comprehensive risk and survival analysis for both individual proteins and the Ang-2/1 ratio. In our studies, we found that Ang-2 itself was not significantly associated with an increased risk of early mortality. It is plausible that Ang-2 might show significance with more statistical power, but it also could be a signal that Ang-2 may play a secondary role compared to Ang-1 in the dysregulation of these proteins and their association with mortality. This approach enhances future research by offering insights into which mechanisms may be more amenable to pharmacological modulation. Our findings suggest that low Ang-1 may have a greater impact on critically ill patients with burn injuries outcomes regarding the dysregulation of TIE2 signaling, prompting us to propose modulation of response with an Ang-1 analog or a TIE2 agonist as our next step in a murine model.

The predictive value of Ang-1 and Ang-2 for mortality after burn injury may stem from their opposing effects on the TIE2 receptor. Ang-1, primarily released by vascular smooth muscle cells, pericytes, platelets, and mesenchymal cells,^26,27 acts as a TIE2 agonist, promoting stable endothelial junctions, anti-inflammatory signaling, and anticoagulant effects.²⁸ In this way, Ang-1 supports oxygen delivery, maintains endothelial integrity, and limits vascular leak, all of which are critical for survival in patients with burn injuries. In contrast, Ang-2 functions as a TIE2 antagonist. Release of Ang-2 results in inhibition of TIE2 signaling and, perhaps most importantly for patients with burn injuries, increases vascular leak and edema formation. It is possible that direct injury in the form of cutaneous burn to the dermal capillary and subcutaneous capillary networks induces endothelial injury and inflammation. This could potentially feed a cycle of increased Ang-2 release and decreased Ang-1 release, collectively resulting in inhibition of TIE2 signaling. Although our study did not measure TIE2 signaling directly, these well-established interactions provide a biological framework to interpret our findings. Increased Ang-2 and decreased Ang-1 may contribute to the endothelial dysfunction and fluid shifts characteristic of burn shock, leading to worse outcomes.^29–31 This study does not identify the source of Ang-1 and Ang-2 in peripheral blood; however, potential sources may include direct release or inhibition from burned tissues that circulate systemically, systemic endothelial release due to stress-related endotheliopathy as described in SHINE,⁴ or a combination of both local and systemic mechanisms. Both potential sources offer accessible treatments in the form of surgical excision of burned tissues and treatment of burn shock³⁰ and burn endotheliopathy³² with plasma inclusive resuscitation. In addition, there are currently multiple Ang-1 and -2 therapeutics in development and in clinical studies.^33–35 These may represent additional adjuncts to treat burn-induced endothelial dysfunction with activation or inhibition of TIE2 based on the clinical phenotype present.

Secondary outcomes

In terms of our secondary outcomes, our results show that patients with larger burns (TBSA > 20%) were more likely to be found with lower Ang-1 levels (≤4.825 ng/mL) and higher Ang-2 levels (>3.554 ng/mL). The Ang-2/1 ratio was also elevated in patients with larger burns. Higher Ang-1 levels were linked to a reduced need for mechanical ventilation, while higher Ang-2 levels and Ang-2/1 ratios were associated with an increased likelihood of requiring ventilation. Thus, Ang-1 may play a protective role, reducing the need for ventilation, while higher Ang-2 and Ang-2/1 ratios may result in elevated risk of requiring ventilation. Since Ang-1 and Ang-2 are significant biomarkers for mortality, future studies should explore their connection to the leading cause of mortality of patients with burn injuries: multisystem organ failure, with a large portion attributed to infectious complications and sepsis.^36,37 Angiopoietins play immunological roles, as Ang-1 and Ang-2 interact with hematopoietic stem and progenitor cells in the bone marrow, which regulate immune responses.^38,39 Understanding how these proteins contribute to immune dysregulation in patients with burn injuries, potentially increasing the risk of infection, is of beneficial investigative interest.

Limitations

At the time this study was designed, no other study had published results on angiopoietin levels in burn injury, as such this cohort was generated to be used as a power calculation of a larger study looking at functional endothelial-derived proteins. The limited size of our cohort may contribute to some of the limitations in Ang-2 discrimination as above as well as in assessing for comorbidities and confounders such as frailty status and inhalation injury, which may influence the initial inflammatory response or the ability to mount a complete immune response.^40–44 Inhalation injury was excluded from multivariate analyses due to limited cases and mortality events, which led to unstable estimates and risked overfitting. Exploratory comparisons showed no significant differences in Ang-1, Ang-2, or their ratio between patients with and without inhalation injury in expired patients (data not shown), suggesting minimal confounding. Although Ang-1 and Ang-2 dysregulation is linked to lung injury and mortality in the literature,^45,46 murine models with Ang-2 overexpression did not increase mortality in LPS-induced lung injury,⁴⁷ indicating angiopoietin imbalance and mortality are not solely explained by acute lung injury despite the common occurrence of this condition in patients with burn injuries.

CONCLUSION

This study adds to the growing body of literature on burn injury by evaluating the relative impact of Ang-1 and Ang-2, including the Ang-2/1 ratio, on PBD1 as a potential marker for prognostication following resuscitation. This novel approach offers insight into patients’ physiology after resuscitation, enhancing the clinical prognostication process. Future studies will need to address these factors by including larger, multi-center cohorts and accounting for regional care differences. In addition, incorporating viscoelastography to assess the magnitude and type of coagulopathy would provide additional insight into the EGL functional status. While assessing total body fluid received was outside the scope of this study, future work will focus on prospectively evaluating resuscitation fluids and stratifying patients based on Ang-1 and -2 levels. Further investigations will also aim to explore the direct contribution of burn tissue in the release of Ang-1 and -2.

Author contributions: Ryan M. Johnson (Data curation [lead], Formal analysis [equal], Investigation [lead], Supervision [supporting], Validation [equal], Visualization, Writing—original draft, Writing—review & editing [lead]), Abigail Plum (Data curation, Investigation [equal], Validation, Writing—review & editing [supporting]), Kevin E. Galicia (Data curation [supporting], Investigation [equal], Validation, Writing—review & editing [supporting]), Irena B. Helenowski (Data curation, Formal analysis [equal], Validation, Visualization, Writing—review & editing [supporting]), Madison D. Kipp (Data curation, Writing—original draft, Writing—review & editing [supporting]), Mary Grace Murray (Investigation, Validation, Writing—review & editing [supporting]), Richard Gonzalez (Validation, Writing—review & editing [supporting]), Mashkoor A. Choudhry (Funding acquisition, Validation, Writing—review & editing [supporting]), and John C. Kubasiak (Conceptualization [lead], Data curation, Formal analysis [supporting], Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision [lead], Validation [equal], Visualization, Writing—original draft, Writing—review & editing [supporting])

Funding: This work is supported by the National Institute of General Medical Sciences (NIGMS: T32GM008750).

Conflict of interest statement: None declared.