Prognostic indicators in acute pancreatitis

This article was originally presented at the 15th International Course on Therapeutic Endoscopy, hosted by the Wellesley Group, Division of Gastroenterology, St Michael's Hospital, Toronto, Ontario, October 9-12, 2002 Royal Infirmary, Glasgow, Scotland Correspondence: Dr Clement W Imrie, Royal Infirmary, 16 Alexandra Parade, Glasgow, Scotland G31 2ER, United Kingdom. Telephone +0141-211-4293, fax +0141-211-4991 CW Imrie. Prognostic indicators in acute pancreatitis. Can J Gastroenterol 2003;17(5):325-328.

E ver since the pioneering work by John Ranson in 1974, a variety of approaches have been used in an effort to identify, at an early stage, the minority of patients with acute pancreatitis (AP) who are at the most risk for severe complications and death.Although Ranson's original criteria applied to alcohol-related AP (1), the criteria were modified in 1979 for use in patients with gallstone-induced disease (2).Amazingly, the latter have been used by very few clinicians or investigators.The original and modified Ranson criteria are listed in Table 1.The modified Glasgow criteria, in which the original 11 Ranson criteria were reduced to eight, were shown in 1984 to be equally applicable to both of the major causes of AP (3).They are listed in Table 2.
The Acute Physiological and Chronic Health Evaluation (APACHE) II system (Table 3) (4) has been increasingly employed, especially since 1989, as a means of grading the severity of pancreatitis and of other serious medical and surgical conditions.Unfortunately, the very large multinational study of the platelet activating factor antagonist, lexipafant, used this system but included a group of 1508 patients with supposedly severe AP who had a mortality rate of only 8% (5).This study used a minimum APACHE II score of six on admission as an inclusion criterion, but most investigators now believe that a score of eight is more appropriate as a threshold for the identification of severe AP.Ranson (using the correct score for the population in question), Glasgow and APACHE II are all reasonably valuable for group analysis, but are too inaccurate for the individual at-risk patient.
C-reactive protein is a single marker that is very useful in determining the severity of the disease (16)(17)(18)(19)(20).However, because it is an acute phase reactant produced in the liver, it does not peak until approximately 36 h after the onset of symptoms.This limitation is similar to that which plagues the Ranson and Glasgow criteria, which require assessment 48 h after admission.Patients with severe AP almost invariably have a C-reactive protein greater than 150 mg/L, but a skew distribution of results makes this test occasionally unreliable.
Johnson and colleagues (21) at Southampton proposed that obesity be added to the APACHE II scoring system, thus creating the APACHE-O system.Although this combination seems promising, it has not been employed in any major prospective study by another research group.
The Atlanta Criteria, developed in 1992, reflect the fact that major organ compromise at an early stage correlates with the presence of severe AP (22).Moreover, the prognosis is worsened in patients who develop major complications later in the of disease, such as a pseudocyst, infection or pancreatic necrosis, but transient organ failure (less than 36 h) is not associated with hight mortality and morbidity.
Many have stated that the gold standard for severe AP is evidence, at contrast-enhanced computed tomography (CT), of reduced perfusion or infarction of the pancreas (19,(23)(24)(25)(26)(27)(28)(29)(30)(31).This criterion has been used to select patients for many trials of early antibiotic therapy.In general clinical practice, however, this test is rarely performed during the initial 24 h after admission.It is more sensible to perform this test on patients who are seriously ill 48 to 72 h later.
The most important prognostic marker of severe disease is objective evidence of significant organ dysfunction, as reflected by the multiple organ dysfunction (MODS) scoring system (32).Failure of two or more organ systems constitutes MODS.It is important to realize that organ failure can occur very early in the course of acute pancreatitis.
The author's group in Glasgow has recently shown that the dynamics of organ function are very important.Organ failure, or deterioration of organ function, that persists for more than 36 h is associated with mortality rates of 50% to 60%, whereas patients with only transient organ dysfunction rarely die from the disease (33).This concept requires validation by clinical trials.Traditional prognostic indices may be deficient, because they do not address the time-course of organ dysfunction.Employing Marshall or similar scores (32) can more clearly identify the maximal risk patients (Table 4).Some investigators believe that there is an excellent correlation between CT evidence of pancreatic damage and clinical organ failure, but this view is not universally held, and more studies are required.The presence of MODS in the first week of illness may account for up to 50% of all deaths from AP in most prospective studies.On the other hand, most deaths that occur after the first week of illness are associated with the presence of infected necrosis of the pancreas and also MODS.
In conclusion, much work still needs to be done in establishing early prognostic factors for acute pancreatitis.Traditional indices that employ clinical findings and laboratory tests are subject to limitations.Contrast-enhanced CT scanning is also useful, but is generally not undertaken early in the course of disease and may exacerbate renal failure.Although MODS is highly predictive, it would be ideal to identify patients who require intensive therapy before the development of severe dysfunction.Exciting new work on the dynamics of organ dysfunction may offer a lead to this important clinical dilemma.

Imrie
Can J Gastroenterol Vol 17 No 5 May 2003 326

TABLE 3 (continued) APACHE-II system
From Knaus et al (4).(A-a)PO 2 Alveolar-arterial oxygen gradient; ABG Arterial blood gas; F i O 2 Fraction of inspired oxygen; HCO 3-Bicarbonate; P a O 2 Arterial partial pressure of oxygen

TABLE 4 Modified Marshall organ failure score (hepatic index excluded)
Data from reference 32.F i O 2 Fraction of inspired oxygen; PO 2 Pressure of oxygen