Selection of Therapeutic Tactics in Pancreatic Necrosis

The dissertation is devoted to the development of differentiated therapeutic tactics for patients with pancreatic necrosis on the basis of systems statistical analysis of the outcomes of complex treatment. It addresses the socioeconomic significance of the disease, the unsatisfactory outcomes of existing therapy, and the methodological difficulties in evaluating the effectiveness of therapeutic interventions. The limitations of univariate statistical approaches are analyzed and the use of multivariate models for predicting disease outcomes depending on the combined influence of conservative and surgical methods is substantiated.

The study is based on retrospective and prospective evaluation of clinical observations using the APACHE II scale for stratification of patients by severity of condition. Inclusion criteria, principles of grouping and randomization, methods of severity assessment, and evaluation of treatment outcomes in the main and control groups have been developed. Original methods for processing acid-base status parameters using the median and interquartile range are proposed. A treatment and diagnostic algorithm has been formulated that takes into account the differentiated influence of modern methods of conservative therapy and surgical interventions on outcomes in patients with prognostically mild and severe forms of the disease.

The rationale for conducting this study was constituted by three sets of problems in the treatment of patients with pancreatic necrosis:

1. high social significance of the disease;

2. unsatisfactory treatment outcomes;

3. the difficulty of adequately evaluating treatment outcomes.

The increasing incidence of acute necrotizing pancreatitis over the past decade (V.K. Gostishchev, V.A. Glushko, 2003; S.F. Bagnenko et al., 2004; T.V. Popov et al., 2008), the relatively young and working-age population of most patients, from 25 to 60 years (V.S. Savelyev et al., 1997), the duration of treatment and the substantial economic costs of its implementation (N.A. Kuznetsov et al., 2005), and the high mortality rates, reaching 20-45% according to V.S. Savelyev et al. (2001), A.S. Yermolov et al. (2003), V.L. Averkiev et al. (2003), E.Yu. Popova et al. (2004), Yu.V. Luzganov et al. (2005), B.S. Briskin, M.D. Dibirov et al. (2007), V.A. Gorsky et al. (2010), R. Isenmann et al. (2001), A. Buter et al. (2002), H.G. Beger et al. (2003), C.D. Johnson, M. Abu-Hilal (2004), all attest to the high social, medical, and economic significance of the disease under study.

The literature contains data on low and even very low mortality rates of 2-13% and purulent complication rates of 7%-14% (S.F. Bagnenko et al., 2004; N.M. Chen et al., 2000; H. Paran et al., 2000; D.A. O'Reilly, A.N. Kingsnoth, 2004). However, the level of evidence in most of these studies, the sample sizes, the assessment of severity of acute pancreatitis, and the methods of statistical analysis raise doubts about their accuracy.

Improvements in the treatment outcomes of patients with pancreatic necrosis in recent decades have been reported by the most authoritative pancreatological surgical forums and in literature reviews based on randomized clinical trials and meta-analyses (V.S. Savelyev et al., 2000; G.S. Rybakov, M.D. Dibirov et al., 2008; B. Gloor et al., 2002; W. Uhl et al., 2002; Working Party of the British Society of Gastroenterology et al., 2005; T. Takada et al., 2006; H.G. Beger, B.M. Rau, 2007). However, many authors emphasize that mortality among patients with severe forms of the disease has decreased only slightly and remains high, at 20%-50%.

Over the past 10-15 years in the Russian Federation, approaches to the complex treatment of patients with necrotizing pancreatitis have undergone substantial changes (V.S. Savelyev et al., 2002; V.K. Gostishchev, V.A. Glushko, 2003; V.L. Averkiev et al., 2003; N.A. Kuznetsov et al., 2005). The principles of antibacterial, antisecretory, and antienzyme therapy, as well as nutritional support, have changed fundamentally. New minimally invasive technologies for percutaneous drainage of fluid parapancreatic collections are being introduced, and surgical tactics are shifting toward limiting early open operations and expanding indications for laparoscopic interventions. These changes are reflected in the recommendations contained in the proceedings of the 9th All-Russian Congress of Surgeons (2000) and the "Protocols for the Diagnosis and Treatment of Acute Pancreatitis" formulated at the Dzhanelidze Research Institute of Emergency Medicine (2004).

Similar changes in the complex treatment of patients with pancreatic necrosis are documented in the materials of the most authoritative foreign surgical pancreatological societies (Working Party of the British Society of Gastroenterology et al., 2005; W. Uhl et al., 2002; T. Takada et al., 2006; M. Hirota et al., 2006; M. Koizumi et al., 2006; K. Takeda et al., 2006; S. Isaji et al., 2006; Y. Kimura et al., 2006; T. Mayumi et al., 2006).

In view of this, it can be stated that the significant changes that have occurred in the complex treatment of patients with pancreatic necrosis have not led to a notable improvement in disease outcomes. This circumstance attests to the complexity of pathogenesis and the severity of the pathology under study and justifies the need for scientific analysis whose main objective should be to identify the methods most significantly influencing improvement of treatment outcomes. Such an analysis is all the more necessary because the published data on the effectiveness of conservative and surgical methods are contradictory.

However, such research is complicated by the fact that patients with pancreatic necrosis are extremely variable in the manifestations of the pathological process and the severity of their condition from the very onset of the disease. Clearly, this requires a differentiated approach to analyzing the effectiveness of treatment methods in patients with differing initial severity.

According to the literature, the principal tool used by researchers to demonstrate the superiority of a particular method or parameter of treatment is the randomized clinical trial (Z. Feng et al., 2001; W. Rosenberger, J.M. Lachin, 2002). A fundamental aspect of these studies is the balancing of study groups with respect to all parameters potentially influencing final treatment outcomes (severity parameters and treatment parameters) except for the one under investigation, achieved through equal probability of patient allocation to comparison groups (O.Yu. Rebrova, 2006; V.W. Berger et al., 2002; B. Dawson, G. Trapp, 2004; A. Biswas et al., 2002). Undoubtedly, this methodology makes it possible to identify correlations between the parameter under study and treatment outcomes (mortality, the frequency of purulent complications, etc.). Moreover, this assessment is most often performed by simple methods of elementary statistical analysis (S. Glantz, 1999; O.Yu. Rebrova, 2006).

However, the difficulties of conducting such studies are also evident. The complex treatment of patients with pancreatic necrosis is intricate and multicomponent, and the severity of patients' condition at the start of treatment is variable; therefore, it is difficult to imagine that groups could be randomized in such a way as to eliminate differences in the influence on treatment outcomes of all possible predictors except one. For this reason, many authors of randomized trials do not provide data on how comparably the complex treatment was administered to patients in the groups under comparison.

Furthermore, to assess the influence of all possible predictors on the final treatment outcomes of patients with pancreatic necrosis, an enormous number of such studies would be required, each of which is economically costly and may last more than one year.

The most serious problem of the studies in question lies above all in an inadequate model of statistical analysis. Univariate statistical methods only allow one to confirm the fact that the groups differ significantly with respect to the compared characteristics (treatment methods and outcomes). This model, at best, permits the assumption of an association (correlation) between these phenomena. However, it does not make it possible to determine the influence of a method on the final outcome, let alone the combined influence of methods, whereas this is precisely the goal of most studies in medicine and pancreatology in particular.

The indicated difficulties can be overcome through a systems approach to analysis. It entails the construction of statistical models for predicting disease outcomes depending on the combined influence of conservative and surgical interventions and dictates the need to distinguish from the overall population of patients (the system) groups of patients (subsystems) that may require different treatment approaches. These research methods are practically not used in surgical pancreatology, evidently because of their complexity and the need to work with a professional statistician. Nevertheless, the possibilities for adequate assessment and solution of the tasks at hand exceed those afforded by univariate statistical analysis methods (B. Dawson, G. Trapp, 2004; A. Biswas, 2008; E.T. Lee, J.W. Wang, 2003).

The totality of the above problems in the treatment of patients with acute necrotizing pancreatitis constituted the rationale for conducting this study.

The objective of the study is to develop, on the basis of systems analysis, a differentiated and scientifically grounded approach to the treatment of patients with pancreatic necrosis that makes it possible to improve the outcome of the disease.

Research tasks

1. To assess the clinical effectiveness of modern methods of treatment of patients with pancreatic necrosis.

2. To determine the effectiveness of modern methods of treatment of patients with pancreatic necrosis in cases with a predicted mild course of the disease (less than 9 points on the APACHE II scale).

3. To determine the effectiveness of modern methods of treatment of patients with pancreatic necrosis in cases with a predicted severe course of the disease (9 points or more on the APACHE II scale).

4. To identify the conservative treatment methods most significantly influencing the reduction of mortality in patients with prognostically mild and severe courses of necrotizing pancreatitis.

5. To determine the surgical treatment methods with a priority influence on the reduction of mortality in patients with prognostically mild and severe courses of necrotizing pancreatitis in the sterile stage of the disease.

6. To determine the surgical treatment methods most significantly influencing the reduction of mortality in patients with prognostically mild and severe courses of necrotizing pancreatitis in the stage of purulent complications.

7. To determine the factors of increased probability of complicated course and lethal outcome in patients with a predicted mild course of pancreatic necrosis.

8. To develop treatment tactics in the presence of risk factors for increased probability of lethal outcome in patients with a prognostically mild course of pancreatic necrosis.

9. To create a treatment and diagnostic algorithm for patients with pancreatic necrosis.

Scientific novelty

1. For the first time, on the basis of systems analysis, it has been established that, depending on the severity of condition according to the APACHE II scale, patients with pancreatic necrosis respond differentially to therapeutic interventions. Assessment of the results of treatment methods in the overall population of patients with pancreatic necrosis is not entirely correct.

2. It has been revealed that modern principles of conservative treatment are most effective in patients with a predicted severe course of pancreatic necrosis, in whom they statistically significantly reduce mortality and alter the survival function by increasing survival time.

3. Using methods of multivariate statistical analysis, it has been revealed that improvement of treatment outcomes in patients with a predicted severe course of pancreatic necrosis is most influenced by the combined action of the following conservative methods: pancreatotropic antibiotics in the stage of sterile necrosis, antisecretory therapy, early nutritional support, HBO therapy, and extracorporeal detoxification. The use of antienzyme therapy, the variants of antiprotease drugs, the timing of their administration from the onset of the disease, and the duration of use do not exert a significant influence on the outcome of the disease.

4. Early sanitation laparotomies without radical necrectomy of the pancreas in the presence of a clinical picture of enzymatic peritonitis, in comparison with laparoscopic sanitation, do not significantly worsen the disease outcome in patients with either an initially severe or mild course of pancreatic necrosis.

5. In purulent complications, minimally invasive ultrasound-guided percutaneous drainages can be effective only in localized fluid collections. In cases of extensive purulent-necrotic retroperitoneal processes and the ineffectiveness of ultrasound-guided drainage of localized infected collections, the practically exclusive alternative is an open intervention with adequate necrectomy and drainage of the retroperitoneal space by a closed or open method.

6. For the first time, by means of statistical modeling, factors of increased probability of lethal outcome have been established in patients with a prognostically mild course of pancreatic necrosis. These are: body temperature below 36.6 °C, leukocytosis exceeding 18 × 10⁹/L, lymphopenia below 10%, and hypoproteinemia below 60 g/L.

7. By methods of systems analysis, features of conservative tactics of complex treatment of patients with pancreatic necrosis have been determined. In cases with a predicted severe course of pancreatic necrosis, as well as a mild course with identified risk factors for high probability of lethal outcome, the treatment complex should include pancreatotropic antibiotic prophylaxis, antisecretory therapy, early nutritional support, HBO therapy, and extracorporeal detoxification. In the absence of these factors in patients with a prognostically mild course, early-from the onset of the disease-prescription of antibiotic prophylaxis and antisecretory therapy is necessary. The choice of drug in this case does not play a decisive role.

Practical significance of the work

1. Conservative treatment methods most substantially improving the outcome of the disease in patients with a predicted severe course of necrotizing pancreatitis have been identified. These are antibiotic prophylaxis with carbapenems and fluoroquinolones, antisecretory therapy with octreotide or quamatel, early nutritional support with parenteral or mixed feeding, HBO therapy, and extracorporeal detoxification.

2. Factors of conservative treatment substantially improving the results of treatment of patients with a predicted mild course of necrotizing pancreatitis have been identified, namely the early-from the onset of the disease-prescription of antibiotic prophylaxis and antisecretory therapy.

3. Tactical approaches to surgical methods of treatment of enzymatic peritonitis have been formulated. In such cases, laparoscopic sanitation is indicated. When it cannot be performed or when diagnostic accuracy is insufficient, sanitation laparotomy without radical removal of necrotic tissues is necessary.

4. Indications for performing minimally invasive ultrasound-guided drainages have been determined. They should be carried out in cases of localized infected fluid processes. In the event of ineffectiveness or impossibility of performing percutaneous drainage, open intervention with adequate necrectomy and sanitation of the focus is indicated.

5. Tactics of surgical treatment for extensive necrotic processes have been formulated: open intervention with adequate necrectomy and drainage of the retroperitoneal space by an open or closed method is necessary.

6. Risk factors of increased probability of lethal outcome in patients with a prognostically mild course of pancreatic necrosis have been determined. These are: body temperature below 36.6 °C, leukocytosis exceeding 18 × 10⁹/L, lymphopenia below 10%, and hypoproteinemia below 60 g/L.

Main propositions submitted for defense

1. Assessment of the effectiveness of modern treatment principles, conducted in the total population of patients with necrotizing pancreatitis, reveals a significant improvement in indices of organ dysfunction according to the APACHE II scale in the first two weeks from the start of treatment, and a change in the survival function with an increase in patient survival time. However, there is no reduction in the number of purulent complications, the frequency of development of multiple organ dysfunction syndrome, or mortality.

2. In the treatment of mild forms of pancreatic necrosis (less than 9 points on the APACHE II scale), modern treatment methods do not contribute to a significant reduction in the frequency of development of multiple organ dysfunction syndrome, purulent complications, or a change in the survival function with an increase in patient survival time. The observed reduction in mortality from 16% to 13% is also not statistically significant.

3. Modern methods of treatment are most effective in the category of patients with a predicted severe course of the disease (9 or more points on the APACHE II scale), in whom they significantly reduce mortality from 72% to 50% and alter the survival function with an increase in patient survival time.

4. The final treatment outcomes of patients with a predicted severe course of pancreatic necrosis are most significantly influenced by the combined prescription of the following conservative treatment methods: pancreatotropic antibiotic prophylaxis with carbapenems or fluoroquinolones, antisecretory therapy, early nutritional support, HBO therapy, and extracorporeal detoxification. In patients with a predicted mild course of the disease, the greatest influence on the disease outcome is exerted by the early-from the onset of the disease-prescription of antibacterial prophylaxis and antisecretory therapy. The choice of drugs in this case is of no decisive significance.

5. In the sterile stage of the disease, in prognostically mild and severe patients with a clinical picture of enzymatic peritonitis, early open sanitation surgical interventions without radical necrectomy and sanitation laparoscopies show no differences in their influence on the disease outcome. In purulent complications, minimally invasive ultrasound-guided drainages can be effective only in localized fluid collections and are therefore more often applicable in prognostically mild cases. In cases of extensive purulent-necrotic retroperitoneal processes and the ineffectiveness of ultrasound-guided drainage of localized infected collections, the exclusive alternative is open intervention with adequate necrectomy and drainage of the retroperitoneal space by a closed or open method. These methods allow 38.3% of patients with purulent complications and a prognostically severe course of the disease to survive (33.3% in purulent-necrotic phlegmons, 54.5% in abscesses) and 68.4% of patients with a prognostically mild course of the disease (52.2% in purulent-necrotic phlegmons, 78.4% in abscesses).

6. In patients with a predicted mild course of pancreatic necrosis, the probability of lethal outcome increases in the presence of the following factors: body temperature below 36.6 °C, leukocytosis exceeding 18 × 10⁹/L, lymphopenia below 10%, and hypoproteinemia below 60 g/L.

7. In the complex treatment of patients with a predicted severe course of necrotizing pancreatitis, antibiotics of the carbapenem or fluoroquinolone groups, antisecretory therapy with octreotide or quamatel, early nutritional support parenterally and enterally, HBO therapy, and extracorporeal detoxification must be prescribed, together with surgical treatment methods when indicated. In the treatment of patients with predicted mild forms of the disease, the use of these methods must be mandatory in the presence of risk factors for increased probability of lethal outcome. In the absence of these factors, early-from the onset of the disease-prescription of antibiotic prophylaxis regardless of the pancreatotropic properties of the drugs, and the prescription of antisecretory therapy are necessary, as are, when indicated, surgical methods.

Implementation of research results into practical healthcare

1. The treatment and diagnostic algorithm for the complex treatment of patients with pancreatic necrosis depending on the severity of their condition is used in the work of the 8th and 19th surgical departments of the Yaroslavl Regional Clinical Hospital, the surgical departments of City Hospital No. 1 of the city of Rybinsk, and the Central District Hospital of the city of Uglich.

2. On the basis of this algorithm, primary triage of patients in healthcare facilities of the region is performed by staff of the Disaster Medicine Center, which makes it possible to hospitalize patients with severe forms of the disease in the regional center in a timely manner and to successfully treat patients with a milder course in district hospitals.

3. The obtained research results are used in the educational process at the Department of Surgery of the Faculty of Postgraduate Education.

Approbation of the work

The main propositions of the work were presented and discussed at the following conferences, meetings, and congresses:

1. Jubilee conference "Modern Technologies of Work of a Multidisciplinary Children's Hospital," Lipetsk, 2003.

2. Second Congress of Anesthesiologists and Resuscitologists of the Central Federal District, Moscow, 2005.

3. Second Scientific Conference "Systems Analysis in Medicine," Blagoveshchensk, 2008.

4. Fiftieth Scientific and Practical Conference dedicated to the 60th anniversary of the Yaroslavl Regional Clinical Hospital, Yaroslavl, 2008.

5. Scientific and Practical Conference of Surgeons of the Central Federal District of the Russian Federation, Yaroslavl, 2008.

6. IX All-Russian Scientific and Practical Conference "Current Issues in the Clinic, Diagnosis, and Treatment of Patients in a Multidisciplinary Medical Institution," St. Petersburg, 2009.

7. Ninth Interregional Conference with International Participation "Parenteral and Enteral Nutrition," Petrozavodsk, 2009.

Publications: Based on the results of the study, 20 printed works have been published. Of these, 10 are in central press, in publications recommended by the Higher Attestation Commission for the publication of works on the topic of doctoral dissertations.