Blood microcirculation indicators in prediction of complicated course of acute pancreatitis

Relevance

Acute pancreatitis is a disease characterized by a phasic course and severe consequences, including pancreatic necrosis. This condition can induce systemic disturbances and inflammatory reactions leading to various complications [1, 2]. It is important to note that cell necrosis is a common phenomenon for all organs and tissues, but it is precisely in pancreatitis that it causes particularly serious alterations in the organism [3, 4].

Within the first 36 hours from disease onset, pancreatic necrosis develops, and further deterioration of the patient's condition is associated not so much with the necrotized areas of the gland, but rather with the development of complications affecting the cardiovascular and respiratory systems, renal function, hepatic insufficiency, metabolic processes, and hemocoagulation [5]. The priority task is to identify among patients with acute pancreatitis those in whom the development of complications is most likely [3]. At the time of admission, the severity of the patient's condition, as well as the results of instrumental and laboratory examination methods, do not allow for an unambiguous conclusion regarding the disease prognosis. At the same time, it has been proven that immediately initiated adequate intensive therapy in the intensive care unit significantly reduces the complication rate and mortality in pancreatitis patients [6].

Objective

To objectify the risk of an aggressive disease course, laser Doppler flowmetry (LDF) is proposed as a method that may assist in predicting destructive complications of acute pancreatitis.

Materials and Methods

The results of blood microcirculation parameter measurements were analyzed in 55 healthy volunteers (control group) and 118 patients treated at the surgical department of the N.A. Semashko Clinical Hospital in Yaroslavl for acute pancreatitis. According to the Atlanta criteria, 58 patients had a mild course of the disease, while the remaining 60 had a severe form. According to the classification by Saveliev V.S. et al. [7], the first group corresponded to abortive pancreatic necrosis, while the second group included patients with fatty and hemorrhagic variants of the pathology. The mean age of the subjects was 54±16.6 years. The healthy volunteer group did not differ from the patient group in terms of sex and age (p>0.05). All subjects provided informed voluntary consent to participate in the study.

The study was conducted in the emergency diagnostic department at a room temperature of 21–24°C. LDF was performed for 8 minutes on the anterior surface of the middle third of the left forearm with the patient in the supine position. The following parameters were measured: blood microcirculation index (MI), standard deviation and coefficient of variation of basal blood flow (σ and Kv, respectively), nutritional blood flow level (M nutr), and normalized amplitudes of endothelial (Ae), neurogenic (An), myogenic (Am), respiratory (Ar), and cardiac (Ac) oscillation ranges. Amplitude normalization was performed relative to 3σ.

The results were processed using the STATISTICA 10 software package. Comparison of groups by quantitative parameters was performed using the Mann–Whitney test. Comparison of frequency characteristics by binary parameters was performed using the χ² test. Statistical significance of differences was confirmed at p<0.05. Predictive capabilities of microcirculation parameters were assessed using ROC analysis, calculating the area under the curve (AUC).

Results and Discussion

In both forms of acute pancreatitis at the time of admission, a decrease in MI on the forearm skin was observed. In the edematous form, its value was 4.51±0.38 perfusion units (p.u.), while in the destructive form—3.42±0.25 p.u. (in both cases, p<0.05 compared to normal). Even more pronounced was the negative dynamics of the nutritional component of skin perfusion: in the destructive form of inflammation, it decreased by a factor of 2 compared to normal—to 1.21±0.15 p.u. In the uncomplicated variant of the disease, M nutr on the forearm was 37.5% below normal (p<0.05). Analysis of the amplitude-frequency spectrum revealed a decrease in normalized amplitudes of endothelial, neurogenic, and myogenic oscillations in both variants of pancreatitis. However, in the absence of destruction, this dynamics was statistically significant only for neurogenic oscillations—a 25.5% decrease (p<0.05). In the complicated course of the disease, the amplitude of endothelial oscillations was more than 2 times below normal; oscillations of the neurogenic and myogenic ranges were below reference values by 45.4% and 36.6%, respectively (p<0.05). Respiratory oscillations at the time of patient admission did not differ from normal in amplitude in both patient groups. The amplitudes of the cardiac oscillation range demonstrated multidirectional dynamics: in edematous pancreatitis, they were elevated compared to normal, averaging 21.89±7.99 p.u. (p>0.05); in destructive pancreatic inflammation, conversely, they decreased to 11.33±1.92 p.u. (p<0.05).

ROC analysis revealed pronounced predictive capabilities for predicting complicated pancreatitis course for the normalized amplitudes of neurogenic and cardiac oscillations. The area under the curve (AUC) for An was 0.857, σ=0.112, 95% confidence interval (CI)=0.558–0.984, p=0.0015. For Ac, AUC=0.889 was obtained; σ=0.115, 95% CI=0.623–0.990, p=0.0007.

For clinical validation of the obtained prognostic system, clinical treatment outcomes of patients with pancreatic necrosis were analyzed. Upon admission, they had no signs of pancreatogenic shock or hemodynamic instability. During LDF performed in the emergency diagnostic department, all patients showed a statistically significant decrease in An and Ac compared to normal. Based on the obtained results, 40 patients were admitted to the intensive care unit (ICU), and from the first hours of hospitalization they received extended syndrome-based treatment including cardiac monitoring, tissue acid-base status control, and daily laboratory monitoring of biochemical markers of pancreatic dysfunction. The comparison group comprised 20 patients who, despite alarming LDF parameter dynamics, were initially admitted to the general surgical ward due to stable hemodynamics, where they received standard infusion therapy with antispasmodic and antiproteolytic agents and further dynamic assessment. Due to the development of pancreatic necrosis clinical manifestations, they were transferred to the ICU.

Fatal outcomes in the main group were recorded in 2 (5%) cases, with 1 patient dying after surgery. In the comparison group, 7 (35%) patients died, of whom 4 died after surgical intervention. A statistically significant difference was obtained regarding the frequency of unfavorable treatment outcomes (χ²=25.93, df=1, p=0.00001). In 2 (10%) patients of the comparison group, acute ischemic cerebrovascular accident developed during hospitalization, requiring transfer to the neurology department. This can be regarded as an additional criterion of unfavorable pancreatic necrosis course, considering the resulting tendency toward hypercoagulation. No acute vascular events were detected in the main group. During the first year after discharge, 4 patients (20%) of the comparison group were re-hospitalized to the surgical department on an emergency basis with clinical manifestations of acute pancreatitis. No readmissions were recorded in the main group during the first year of long-term follow-up.

When recording an LDF-gram, one can measure the microcirculation index, its standard deviation, and coefficient of variation, which at the stage of the previous generation of stationary laser microcirculation analyzers formed the basis for diagnosing skin perfusion [8], as well as the variable components of microblood flow—amplitudes of microcirculation control factors. Located in different frequency ranges of oscillations, they reflect the influence of specific active and passive mechanisms regulating the tone and filling of microcirculatory unit vessels. From a pathogenetic perspective, changes in the amplitudes of the cardiac and neurogenic ranges are of particular interest. The amplitude of cardiac oscillations depends on the inflow of blood into the microcirculation system with the pulse wave; already at the initial stages of decreased myocardial contractile activity, it will show negative dynamics. A decrease in the magnitude of neurogenic oscillation amplitude may be a sign of increased sympathetic influence on arterioles [9]. Preliminary studies conducted on this matter indicate the prognostic significance of these changes in the complicated course of acute pancreatitis [10]. Our data confirm that the maximum impairment of the hemomicrocirculation system jeopardizes the severe course of pancreatic necrosis in both abortive and complicated variants of the disease.

Conclusion

In patients with acute pancreatitis, laser Doppler flowmetry allows the identification of several parameters with high predictive potential for complicated disease course already at the admission stage. The most significant are the normalized amplitudes of neurogenic and cardiac oscillations (AUC=0.857 and AUC=0.889, respectively). Consideration of laser Doppler flowmetry results when making decisions regarding ICU admission can reduce overall and postoperative mortality in destructive pancreatitis and improve clinical treatment outcomes.