- Citado por SciELO
versión impresa ISSN 1130-0108
Rev. esp. enferm. dig. v.96 n.10 Madrid oct. 2004
Early onset of organ failure is the best predictor of mortality in acute pancreatitis
I. Poves Prim, J. Fabregat Prous, F. J. García Borobia, R. Jorba Martí, J. Figueras Felip and E. Jaurrieta Mas
Department of General and Gastrointestinal Surgery. Hospital Prínceps d'Espanya. Citutat Sanitària i Universitària de Bellvitge.
L'Hospitalet de Llobregat. Barcelona. Spain
Background: APACHE II is a multifactorial scoring system for predicting severity in acute pancreatitis (AP). Organ failure (OF) has been correlated with mortality in AP.
Objectives: to evaluate the usefulness of APACHE II as an early predictor of severity in AP, its correlation with OF, and the relevance of an early establishment of OF during the course of AP.
Patients and methods: from January 1999 to November 2001, 447 consecutive cases of AP were studied. APACHE II scores and Atlanta criteria were used for defining severity and OF.
Results: twenty-five percent of patients had severe acute pancreatitis (SAP). APACHE II at 24 h after admission showed a sensitivity, specificity, and positive and negative predictive value of 52, 77, 46, and 84%, respectively, for predicting severity. Mortality for SAP was 20.5%. Seventy percent of patients who developed OF did so within the first 24 hours of admission, and their mortality was 52%. Mortality was statistically significant (p < 0.01) if OF was established within the first 24 hours after admission.
Conclusions: APACHE II is not reliable for predicting outcome within the first 24 hours after admission and should therefore be used together with other methods. OF mostly develops within the first days after admission, if ever. The time of onset of OF is the most accurate and reliable method for predicting death risk in AP.
Key words: Severe acute pancreatitis. Organ failure. APACHE II.
Poves Prim I, Fabregat Prous J, García Borobia FJ, Jorba Martí R, Figueras Felip J, Jaurrieta Mas E. Early onset of organ failure is the best predictor of mortality in acute pancreatitis. Rev Esp Enferm Dig 2004; 96: 705-713.
Correspondencia: Ignasi Poves. Servicio de Cirugía General y Digestiva. Ciutat Sanitària i Universitaria de Bellvitge. C/Feixa Llarga, s/n.08907 L'Hospitalet de Llobregat. Barcelona. e-mail: firstname.lastname@example.org
The APACHE II scoring system (1) is the best illness grading system for quantifying severity in acute diseases. It is relatively easy to perform and its usefulness has been proved in the management of acute pancreatitis (AP) (2). In contrast to other scoring systems, it may be used since patient admission and during the whole course of disease. A good correlation between APACHE II scores and clinical outcome has been demonstrated in AP (3,4).
The International Symposium on AP held in Atlanta in 1992 promoted the use of the APACHE II scoring system, or other such as Ranson's, to evaluate disease severity. AP was considered severe if organ failure (OF) and/or local complications such as necrosis, abscess or pseudocyst were associated (5). Severe acute pancreatitis (SAP) was further characterized by eight or more points in the APACHE II score or three or more Ranson criteria.
This study aimed to assess the true usefulness of APACHE II as an early predictor of severity in patients with AP, the correlation of APACHE II scores with outcome and OF onset, and the relationship between mortality and early onset of OF in patients with SAP.
PATIENTS AND METHODS
Between January 1999 and November 2001 we studied 447 consecutive patients admitted to our hospital with a diagnosis of AP. This diagnosis was based on the presence of compatible clinical features and symptoms, and a serum amylase level over three times above normal (normal < 2.16 ukat/l), and/or a urine amylase level greater than 20 ukat/l (normal < 18 ukat/l). Patients with a doubtful diagnosis and those whose initial diagnosis of pancreatitis was modified because of imaging studies were excluded. Patients with chronic pancreatitis were also excluded.
The APACHE II score was calculated at admission, at 24 h, at 48 h, and then daily until clinical improvement or death. Physiological data (temperature, heart rate, blood pressure, and respiratory rate) were recorded at least every 6 hours, and the Glasgow Coma Scale was calculated daily. The APACHE II score was determined by using the most abnormal physiological and laboratory measurements during each 24 h period. In order to study early organ failure, patients were included if admitted within 48 h of symptom development.
A contrast-enhanced computed tomogram (CT) was performed if local complications were suspected (abdominal mass or tenderness, persistent abdominal pain, etc.), in the presence of clinical deterioration or OF, or when the diagnosis of AP was doubtful. Blood cultures and a CT-guided percutaneous fine-needle aspiration (PFNA) were performed when pancreatic infection was suspected (high fever and leukocytosis, OF or clinical deterioration).
The Atlanta classification system, which is based on clinical symptoms and radiological findings, was used to assess the severity of AP (5,6). AP was considered severe when associated with OF (Table I) and/or local complications such as pancreatic necrosis, abscess, and large or multiple acute fluid collections. Necrosis was diagnosed by CT when involving at least 30% of the pancreatic gland. Patients with initial hypoxia, transient hypotension or acute renal failure at admission -attributed to previous dehydration- who responded rapidly to therapy (fluid resuscitation) were not considered severe cases if pancreatic necrosis was not found on abdominal CT. A CT scan was performed at 72 hours after admission if the response to medical treatment was good. In those with a poor outcome a CT scan was carried out within the first 48 hours.
Surgical treatment was initially only indicated for patients with infected pancreatitis diagnosed by PFNA. Patients with sterile pancreatitis were all initially treated by organ dysfunction support, and surgical débridement was performed when intensive medical care was not effective.
The statistical analysis for comparing APACHE II data was performed using Student's t-test. Categorical variables were analyzed using Fisher's two-tailed exact test. A p-value < 0.05 was considered significant for both.
Factors involved in the etiology of AP included: gallstones in 266 patients (59.5%), alcohol abuse in 61 (13.6%), and other recognized factors (endoscopic cholangiopancreatography, drugs, neoplasm, etc.) in 34 (7.6%). No associated factors were identified in 86 patients (19.3%) during hospital stay.
According to the Atlanta classification system, 335 (75%) patients suffered from mild AP and 112 (25%) from SAP. Peak C-reactive protein (CRP) was 89 ± 88 and 277 ± 111 mg/l within the first 48 h after admission for mild and severe AP, respectively. Of the 112 patients classified as having SAP, 57 developed OF, 62 necrotizing pancreatitis, and 100 multiple acute fluid collections.
Thirty-seven patients with SAP needed surgery. Indications for surgical treatment included: sepsis syndrome with positive blood cultures and/or PFNA in 27 patients (73%), sterile pancreatitis with persistent OF or progressive clinical deterioration despite medical therapy in intensive care unit in seven patients (18.9%), massive intraabdominal bleeding in two patients (5.4%), and necrotizing pancreatitis from duodenal-pancreatic reflux caused by an upper intestinal obstruction (bezoar) in one patient (2.7 %).
Twenty patients were operated upon during the first week, eight during the second, four during the third, and five during or beyond the fourth week after admission (Fig. 1). Most patients operated on during the first week underwent surgery after the first 4 days of admission, and 65% had an infected necrotizing pancreatitis.
APACHE II score
In eight patients we were unable to fully calculate a complete APACHE II score at admission and at 24 h because of missing data (patients referred from other centers). Mean APACHE II score at admission and at 24 hours, and peak scores for mild and severe AP were 6.1 ± 3.8, 5.3 ± 3.7, 6.5 ± 4, and 8.6 ± 4.8, 9.1 ± 5.1, 12.2 ± 6.3, respectively. While in mild AP peak scores were reached at 0.3 ± 0.1 days after admission (first 24 h), in severe attacks they were reached at 4.1 ± 7.8 days (after 48 h). Eighty-one percent (270/331) of patients with mild pancreatitis showed peak APACHE II scores at admission, while only 34.2% (37/108) with a severe attack did so. In all, 94% (311/331) and 56.5% (60/108) of patients with mild and severe attacks, respectively, had a peak score within 24 hours of admission.
Thirty-seven percent (126/335) of patients with mild pancreatitis had a peak score ≥ 8. One hundred and one of these were over 65 years of age, and seven had a severe chronic health status. In about 85% of patients with mild AP and APACHE II ≥ 8, this score could be attributed to the additional weighting of age and chronic disease exclusively, not to inflammatory response.
Table II shows the sensitivity, specificity, and positive and negative predictive values of an APACHE II cut-off ≥ 8 for detecting severity depending on the day after admission.
The APACHE II score at admission and at 24 h, and the peak score were 10.4 ± 5.2, 11.3 ± 5.4, and 15.6 ± 6.5 for patients who had SAP with OF. In SAP without OF scores of 6.7 ± 3.4, 6.6 ± 3.4, and 8.4 ± 3.2, respectively, were recorded. These scores were statistically significant between groups depending of the presence of OF (p < 0.01).
No deaths occurred in the mild pancreatitis group, and none of these patients had to be operated on. In contrast, the overall mortality rate was 20.5% (23/112) for SAP. Mortality was 40.3% (23/57) and 0% (0/55) when SAP developed with or without OF, respectively. Causes of mortality included: thirteen patients had persistent multiple organ failure (MOF) not responding to intensive medical treatment and/or surgical therapy (9 sterile and 4 infected pancreatitis); five patients died from intractable acute respiratory distress syndrome; one from intracranial hemorrhage; two from persistent intra-abdominal sepsis; one from intra-abdominal diffuse hemorrhage, and one from upper gastrointestinal massive bleeding. Ten of these deaths occurred in elderly patients (> 75 years of age) with severe pre-existing co-morbidity (seven with severe heart failure and three with severe chronic respiratory failure). Four patients who died had morbid obesity, and one had chronic renal failure.
Nineteen patients presented with OF at admission, twenty-one developed it within the next 24 h, nine between 24 h and 48 h, and 8 after 48 h of admission. Mortality was higher in patients who developed OF within 24 hours of admission versus those whose OF was of later onset (p < 0.01) (Table III).
Twenty-six of 37 operated on patients had preoperative OF. Mortality in these was 46% (12/26). None of the patients operated on without pre-operative OF died.
In sixteen of twenty-three patients who died, pancreatic necrosis was diagnosed on abdominal CT and/or operatively. Mortality was significantly higher in patients with pancreatic necrosis associated, varying from 25.8% (16/62) to 14% (7/50).
AP is a complex disease that may remain localized in the pancreas, spread to regional tissues, or even involve distant organs (6). Microcirculatory disorders can be found not only in the pancreas, but also in the liver, colon, and lung (7,8). AP is a disease with difficult-to-prevent complications in patients identified as at risk (9). The search for early predictors of severity is motivated by a poor response to medical treatments available (10, 11). Supporting vital organs is the only effective medical treatment we know of (12,13). A great number of predictors of severity, such as APACHE II (2-4) and contrast-enhanced CT (12,15,16), have been proposed, but their influence on prognosis improvement is unclear, and the results of SAP therapy are still discouraging.
Given that AP is considered a systemic disease, a number of scoring systems are useful to evaluate severity, as they are all based on physiological and biochemical variables. APACHE II is the multiple-scoring system with the highest accuracy in predicting severity (3,4). Application of this system within the first 24 h may provide as good a prediction of severity as Ranson and Glasgow scores (17), but also a better follow-up throughout the course of disease. Recently, other methods for early prediction of severity have been reported (18). A new APACHE III prognostic system has been suggested to provide better risk stratification in severely ill hospitalized patients (19), but its advantages over APACHE II remain unclear (20). Neither serum polymorphonuclear elastase (21), nor urine trypsinogen activation peptide (TAP) provided accurate severity prediction as single markers 24 hours after symptom onset (22). In a multicenter study by Neoptolemos, the sensitivity, specificity, and positive and negative predictive values of urinary TAP at 24 h were 58, 73, 39, and 86%, respectively (23). These values are similar to those observed in our study for APACHE II at admission and at 24 h. None of the reported early predictive factors are ideal, and all -including APACHE II, as demonstrated in our study- remain imperfect.
Mortality in AP is usually associated with systemic complications arising during the course of disease (10,24). OF is the clinical manifestation of a severe systemic inflammatory response caused by a release of cytokines, free radicals and vasoactive molecules into the circulatory system (25,26). Microcirculatory disorders in the pancreas and remote organ systems extend beyond the early stage of AP and can persist for more than 48 h (6). The mortality rate in patients with necrotizing pancreatitis is higher when OF is associated, and varies from 20 to 40% depending on the various series (27-30). The incidence of OF should be established by bacterial infection and extent of necrotic parenchyma in sterile necrosis (31). Beger et al. showed that pancreatic infection is more common during the late course of disease (32,33). In these patients OF occurs long after admission. When sterile necrotizing pancreatitis is accompanied by OF, this complication usually develops soon after the onset of the disease (27,31).
In our study we found data adding to the notion that the severity of the inflammatory process involving systemic organs in AP is established early during the first hours of disease. As has been proved in other studies (29), there is a subgroup of patients with early SAP, defined as the presence of OF at admission, who more frequently develop intractable OF. Twenty-one of 23 patients who died developed OF within 24 hours or less. This is a very strong statement since the absence of OF within the first 24 hours ends up being a very good predictor of survival. As Buter et al. (34) reported, deteriorating organ dysfunction during the first week is a highly significant determinant of survival; in our opinion, however, the onset timing of OF is also extremely important. The later OF develops, the better the prognosis. Surgical treatment provides good control for local complications occurring in infected necrosis, but has not this same role for controlling systemic complications. Surgical treatment will be more effective if needed later. In order to obtain good results our desire is always to delay surgery when clinical conditions are favorable (supporting therapy, antibiotics and, in selected cases, percutaneous drainage). Unfortunately, this was not possible for most of our patients in need of surgery because of infected SAP or sterile SAP with uncontrollable OF, who had to be operated on within the first week of admission (Table II).
Considering, as the Atlanta consensus conference recommended, APACHE II ≥ 8 points as predictive of severity in AP (5), this score showed a poor sensitivity, specificity, and positive predictive value. Such poor sensitivity was probably due to the added weight of age and chronic heath status. From our results, we consider that an APACHE II score cut-off value ≥ 8 in patients without OF will probably not reflect severity in AP. No deaths were recorded in patients with a score ≥ 8 and without OF; however, all patients with OF who died had a peak score ≥ 8. These results are similar to those reported by Lankisch et al. (35), who concluded that APACHE II scores on admission were unreliable for diagnosing necrotizing pancreatitis; we obtained similar values for sensitivity, specificity, and positive and negative predictive value. Although APACHE II is considered the most precise scoring system for detecting severity in AP to date, and correlates well with SAP, it cannot be used alone. This system should be used together with clinical observations made by an expert physician and an evaluation of OF evidence.
OF is the main prognostic factor for detecting an unfavorable outcome and a high risk of death in AP. An APACHE II cut-off ≥ 8 is not reliable for predicting outcome within the first 48 h of admission, and its main usefulness in AP is to serve as a model of reference for testing new predictive methods, and for comparing patients in inter-institutional series. Patients with SAP who develop OF within the first 24 h of admission are those with a worst prognosis and probably a fatal outcome. A strict follow-up of the parameters defined as OF in the Atlanta consensus within the first 24 h of admission is the most accurate and reliable method for predicting severity in AP, since the risk of mortality is greater than 50%.
1. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13: 818-29. [ Links ]
2. Banks PA. Practice guidelines in acute pancreatitis. Am J Gastroenterol 1997; 92: 377-86. [ Links ]
3. Larvin M, McMahon MJ. APACHE II score for assessment and monitoring of acute pancreatitis. Lancet 1989; II: 201-4. [ Links ]
4. Wilson C, Heads A, Imrie CW. Prediction of outcome in acute pancreatitis: a comparative study of APACHE II, clinical assessment and multiple factor scoring systems. Br J Surg 1990; 77: 1260-4. [ Links ]
5. Bradley III E. A clinically based classification system for acute pancreatitis. Arch Surg 1993; 128: 586-90. [ Links ]
6. Banks PA. A new classfication system for acute pancreatitis. Am J Gastroenterol 1994; 89: 151-2. [ Links ]
7. Foitzik T, Eibl G, Hotz B, Hotz H, Kahrau S, Kasten C, et al. Persistent multiple organ microcirculatory disorders in severe acute pancreatitis: experimental findings and clinical implications. Dig Dis Sci 2002; 47: 130-8. [ Links ]
8. Mozo G, Del Olmo ML, Caro-Patón A, Reyes E, Manzano L, Belmonte A, et al. Lung changes and cytokine levels in a model of experimental acute pancreatitis. Rev Esp Enferm Dig 2002; 94: 53-66. [ Links ]
9. Beger HG, Rau B, Isenmann R. Prevention of severe change in acute pancreatitis: prediction and prevention. J Hepatobiliary Pancreat Surg 2001; 8: 140-7. [ Links ]
10. Banks PA. Acute pancreatitis: medical and surgical management. Am J Gastroentol 1994; 89: S78-85. [ Links ]
11. Wyncoll DL. The management of severe acute necrotizing pancreatitis: an evidence-based review of the literature. Intensive Care Med 1999; 25: 146-56. [ Links ]
12. Banks PA. Predictors of severity in acute pancreatitis. Pancreas 1991; 6 (Supl. 1): S7-S12. [ Links ]
13. Ribeiro MD, Paiva JA, Landeiro N, Duarte J. Patients with severe acute pancreatitis should be more often treated in an Intensive Care Department. Rev Esp Enferm Dig 2002; 94: 523-32. [ Links ]
14. Zhu AJ, Shi JS, Sun XJ. Risk factors influencing mortality of patients with severe acute pancreatitis within 24 hours after admission. Hepatobiliary Pancreat Dis Int 2003; 2: 453-7 [ Links ]
15. Balthazar EJ, Robinson DL, Megibow AJ, Ranson JHC. Acute pancreatitis: value of CT in establishing prognosis. Radiology 1990; 174: 331-6. [ Links ]
16. Basterra G, Álvarez M, Marcaide A, Delgado E, Díaz de Otazu R, García Campos F. Acute pancreatitis: evaluation of the prognostic criteria of the latest Baltasar tomographic classification. Rev Esp Enferm Dig 1999; 91: 433-8. [ Links ]
17. Johnson CD, Toh S. Prediction of severity in acute pancreatitis. In: Pancreatic disease, Towards the year. 2nd ed. Springer, 2000. [ Links ]
18. Yadav D, Agarwal N, Pitchumoni CS. A critical evaluation of laboratory tests in acute pancreatitis Am J Gastroenterol 2002; 97: 1309-18. [ Links ]
19. Knaus WA, Wagner DP, Draper EA, Zimmerman JE, Bergner M, Bastos PG, et al. The APACHE III prognostic system. Risk prediction of hospital mortality for critically ill hospitalized patients. Chest 1991; 100: 1619-36. [ Links ]
20. Chatzicostas C, Roussomoustakaki M, Vlachonikolis IG, Notas G, Mouzas I, Samonakis D, et al. Comparison of Ranson, APACHE II and APACHE III scoring systems in acute pancreatitis. Pancreas 2002; 25: 331-5. [ Links ]
21. Luengo L, Castellote M, Ros S, Feliu F, Vadillo J, Olona C. Clinical usefulness of the determination of the lipase/elastase ratio and of the polymorphonuclear elastase in the hospital admission of the patient with acute pancreatitis. Rev Esp Enferm Dig 1996; 88: 551-4. [ Links ]
22. Tenner S, Fernández-del Castillo C, Warshaw A, Steinberg W, Hermon-Taylor J, Valenzuela JE, et al. Urinary trysinogen activation peptide (TAP) predicts severity in patients with acute pancreatitis. Int J Pancreatol 1997; 21: 105-10. [ Links ]
23. Neoptolemos JP, Kemppainen EA, Mayer JM, Fitzpatrick JM, Raraty MG, Slavin J, et al. Early prediction of severity in acute pancreatitis by urinary trypsinogen activation peptide: a multicentre study. Lancet 2000; 355 (9219): 1955-60. [ Links ]
24. Bradley EL. Necrotizing pancreatitis. Br J Surg 1999; 86: 147-8. [ Links ]
25. Norman J. The role of cytokines in the pathogenesis of acute pancreatitis. Am J Surg 1998; 175: 76-83. [ Links ]
26. Karne S, Gorelick F. Etiopathogenesis of acute pancreatitis. Surg Clin North Am 1999; 79: 699-710. [ Links ]
27. Rau B, Pralle U, Uhl W, Schoenberg MH, Beger HG. Management of sterile necrosis in instances of severe acute pancreatitis. J Am Coll Surg 1995; 181: 279-88. [ Links ]
28. Poves I, Fabregat J, Biondo S, Jorba R, Borobia FG, Llado L, et al. Results of treatment in severe acute pancreatitis. Rev Esp Enferm Dig 2000; 92: 586-90. [ Links ]
29. Isenmann R, Rau B, Beger HG. Early severe acute pancreatitis: characteristics of a new subgroup. Pancreas 2001; 22: 274-8. [ Links ]
30. Soran A, Chelluri L, Lee KK, Tisherman SA. Outcome and quality of life of patients with acute pancreatitis requiring intensive care. J Surg Res 2000; 91: 89-94. [ Links ]
31. Isenmann R, Rau B, Beger HG. Bacterial infection and extent of necrosis are determinants of organ failure in patients with acute necrotizing pancreatitis. Br J Surg 1999; 86: 1020-4. [ Links ]
32. Beger HG, Bittner R, Block S, Büchler M. Bacterial contamination of pancreatic necrosis. A prospective clinical study. Gastroenterology 1986; 91: 433-8. [ Links ]
33. Beger HG, Rau B, Mayer J, Pralle U. Natural course of acute pancreatitis. World J Surg 1997; 21: 130-5. [ Links ]
34. Buter A, Imrie CW, Carter CR, Evans S, McKay CJ. Dynamic nature of early organ dysfunction determines outcome in acute pancreatitis. Br J Surg 2002; 89: 289-302. [ Links ]
35. Lankisch PG, Warnecke B, Bruns D, Werner HM, Grossmann F, Struckmann K, et al. The APACHE II score is unreliable to diagnose necrotizing pancreatitis on admission to hospital. Pancreas 2002; 24: 217-22. [ Links ]