| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Both authors: Department of Radiology, NYU School of Medicine–Bellevue Hospital Center, 462 First Ave., NB 3W33A, New York, NY 10016.
Received June 21, 2007;
accepted after revision September 28, 2007.
Address correspondence to D. K. Lenhart
(dipti.kandlikar{at}med.nyu.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Initial MDCT studies of 169 consecutive patients with mild acute pancreatitis and 203 follow-up CT examinations were reviewed. The fate of peripancreatic fluid collections was investigated, and the incidence and type of local complications were recorded and correlated to the CT grading system (A–E).
RESULTS. Complications developed in nine of 169 patients, for an incidence of 5.3%. All morbidity occurred in the subgroup of 73 patients with initial fluid collections, for an incidence of 12.3%. Follow-up CT examinations available in 51 of these 73 patients documented rapid fluid resolution in 35 cases (68.6%) and persistence of fluid more than 2 weeks from onset in seven asymptomatic patients (13.7%). Acute, life-threatening complications (hemorrhage, infection, perforation) occurred in five patients, for an incidence of 6.8% among the 73 patients with initial fluid collections, or 3.0% in the entire group of 169 patients. Five patients developed acute pseudocysts. Long-term follow-up studies discovered two patients with chronic pancreatitis and one with groove pancreatitis.
CONCLUSION. A small number of acute, life-threatening abdominal complications and chronic complications are expected to occur in patients with milder forms of acute nonnecrotizing pancreatitis presenting with fluid collections. In these patients, clinical monitoring and repeated imaging studies are recommended to document the resolution of fluid or the development of complications.
Keywords: abdominal imaging • acute pancreatitis • MDCT • pancreas
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
According to this classification system, an acute attack of pancreatitis is divided into two major clinical forms: First, mild acute pancreatitis occurring in approximately 80% of patients has no CT evidence of necrosis, exhibits minimal or no distal organ dysfunction, and shows rapid recovery without complications. It is a self-limiting disease previously called "edematous" or "interstitial" pancreatitis. Second, severe acute pancreatitis, also called "necrotizing pancreatitis;" occurs in approximately 20% of patients; shows CT evidence of parenchymal necrosis (lack of enhancement); and exhibits systemic manifestations, distal organ failure, a protracted clinical course, and an increased incidence of morbidity and mortality [6, 7]. Indeed, most patients who develop local complications have necrotizing pancreatitis [8–11]. The mortality incidence is less than 1% in mild pancreatitis, with a striking increase to 10–23% in the presence of pancreatic necrosis [3, 10, 11]. Furthermore, more than 50% of deaths do not occur immediately, but rather within a few weeks after an acute episode secondary to abdominal complications and occur mainly in patients with pancreatic necrosis [1, 11]. This clinical classification emphasizes the importance of pancreatic necrosis as a predictive indicator, while overlooking intermediary forms of disease presenting with fluid collections but without necrosis.
The purpose of this retrospective study was to estimate whether and to what degree milder forms of pancreatitis without necrosis contribute to the development of local abdominal complications. We attempted to determine the fate and outcome of extravasated peripancreatic fluid collections and to assess the incidence and type of local morbidity and patient outcome as correlated with the CT grading scale [12] in patients without pancreatic necrosis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In addition to the 169 initial CT scans obtained within 24 hours of patient presentation to the hospital, we reviewed 203 follow-up CT examinations, for a total of 372 studies or an average of 2.2 examinations per patient. Among the 169 initial episodes, 82 patients (48.5%) had follow-up examinations, for an average of 3.5 scans per patient in this subgroup. The follow-up time ranged from 4 to 880 days, with an average time to final CT examination of 124 days. Seventy-three percent of patients with follow-up studies underwent their first follow-up CT within 60 days of their initial CT examination.
Our series of 169 subjects was composed of 93 males and 76 females with an age range of 11–90 years (average age, 49 years). The cause of pancreatitis was gallstones in 57 patients, alcohol in 44 patients, gallstones and alcohol combined in nine patients, other causes (including hyperlipidemia, lupus, pancreas divisum, and post-ERCP) in 10 patients, and unknown in 49 patients.
CT Technique
Initial CT examinations were performed on a 16-MDCT scanner (Sensation 16,
Siemens Medical Solutions) in 103 patients and on a 4-MDCT scanner
(LightSpeed, GE Healthcare) in 66 patients.
One hundred twenty-five patients (74%) were clinically suspected to have acute pancreatitis and were scanned using our institution's two-phase acquisition pancreatic protocol. These patients were instructed to drink 500 mL of water for negative opacification of the gastrointestinal tract immediately before imaging. The initial pancreatic phase (late arterial dominant phase) of the examination was performed over the upper abdomen from T11 to L3 vertebral body levels with a scanning delay of 40 seconds after the start of IV administration of 1.5 mL/kg of contrast material (300 mg I/mL, Ultravist [iopromide, Bayer HealthCare] or Omnipaque [iohexol, GE Healthcare]) at an injection rate of 4 mL/s. On the 16-MDCT scanner, the images were acquired at 120 kVp with a detector row configuration of 16 x 0.75 mm and a table speed of 9.0 mm per rotation with a reconstructed slice thickness of 3 mm. On the 4-MDCT scanner, the images were acquired at 120 kVp with a detector row configuration of 4 x 1.25 mm and table speed of 7.5 mm per rotation with a reconstructed slice thickness of 2.5 mm.
The second portal-dominant phase of the examination was performed from the diaphragm to the symphysis pubis at an 80-second scanning delay. On the 16-MDCT scanner, the images were acquired at 120 kVp with a detector row configuration of 16 x 1.5 mm, table speed of 18.0 mm per rotation, and reconstructed slice thickness of 4 mm. On the 4-MDCT scanner, the images were acquired at 120 kVp with a detector row configuration of 4 x 2.5 mm, table speed of 15 mm per rotation, and reconstructed slice thickness of 5 mm.
Forty-four patients (26%) were scanned using a single-phase portal venous acquisition with an 80-second scanning delay at an IV contrast injection rate of 3 mL/s, using the same CT scanner parameters as the second phase of the pancreatic protocol. These patients drank 1.5 L of diluted (2%) water-soluble contrast material (Gastrografin [meglumine diatrizoate], Bristol-Myers Squibb) beginning 1 hour before imaging.
Data Analysis
Reconstructed axial images were reviewed on our PACS workstation (either
PACS MagicView 1000, Siemens Medical Solutions; or Philips PACS, Philips
Medical Systems) and thinner collimation and coronal or multiplanar
reformation were performed if needed. All images were reviewed together by two
radiologists, one with 40 years of experience in abdominal imaging and the
other a radiology resident, both of whom were blinded to clinical follow-up
information.
The initial episodes of pancreatitis were stratified into five groups (grades A–E) according to the previously described CT grading scale [12] (Table 1). The follow-up studies were evaluated for the number and type of local complications, including pseudocysts, hemorrhage, infection, bowel perforation, venous thrombosis, pseudoaneurysm, and chronic and groove pancreatitis. The complications were correlated with the initial CT grade. Fluid collections that develop immediately after an episode of pancreatitis due to leakage of pancreatic secretions define grades D and E pancreatitis. These are ill-defined nonencapsulated collections to be distinguished from pseudocysts, which are completely encapsulated fluid collections that develop more than 4 weeks after the initial episode of pancreatitis [6, 7].
|
A two-tailed Fisher's exact test was used to compare the difference in complication rate between each grade of pancreatitis. It was assumed that those patients without follow-up examinations who clinically improved and became asymptomatic did not develop complications. We reviewed our hospital's clinical electronic data repository for follow-up findings, need for surgical intervention, and final clinical outcome.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
0.001); indeed, no complications occurred in those patients without fluid
collections.
|
|
|
|
|
The type and number of local complications in our series of 169 patients with acute pancreatitis as correlated to the A–E grad ing system are presented in Table 4. These complications developed entirely in our patients with fluid collections (grades D and E). Acute short-term life-threatening complications (hemor rhage, infection, or perforation) developed in five of 73 patients with peripancreatic fluid (6.8%), and chronic long-term morbidity (chronic pancreatitis or groove pancreatitis) was seen in three of 73 patients (4.1%).
|
Among the 73 patients with initial fluid collections, we detected five patients who developed single pseudocysts from 2 x 1 cm to 7 x 6 cm, for an incidence of 6.8% (Fig. 3A, 3B, 3C). Two of the pseudocysts developed hemorrhage and rupture with leakage of blood into the peritoneal cavity (Fig. 4A, 4B). One of these patients also developed duodenal perforation and required surgery, and the other recovered without surgical intervention.
|
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Digestion of the pancreatic gland or of peripancreatic tissues after the leakage of activated pancreatic secretions from acinar cells is responsible for the development of local complications [14]. Enzymatic fluid secretions dissect fascial planes and have a deleterious effect on vascular structures, adjacent hollow or solid organs, and retroperitoneal fat, producing fat necrosis. Extensive retroperitoneal fat necrosis interferes with the rapid absorption of extravasated, and sometimes hemorrhagic, fluid collections. When these transitory collections are not rapidly absorbed or continue to increase in size, they tend to organize and loculate by developing partial capsules. Because liquefied necrotic tissue, blood products, and retained fluid are excellent media for bacterial growth, infection may develop [14]. In the initial 1–2 weeks after an acute attack, however, the natural evolution of sterile fluid collections remains unpredictable, so we recommend that these collections be followed up with imaging examinations in symptomatic patients.
In our series of 169 patients, fluid collections were detected in 73 patients (43.2%) and almost totally resolved within 7–10 days in most patients. We were able to document resolution in 35 of the 51 patients (68.6%) in whom follow-up CT studies were available for review (Table 3). The remaining 22 patients in whom long-term follow-up studies were not obtained all had small fluid collections (grade D). These patients had an uneventful clinical course and rapid improvement, and it may be presumed that because these patients remained asymptomatic, their small fluid collections resolved as well.
The occurrence and fate of fluid collections in acute pancreatitis have been previously reported in a series of 48 patients with and without pancreatic necrosis [15]. In that series, fluid was found in 37% of patients. It resolved spontaneously in about half the patients and led to complications (pseudocyst, abscess, infected necrosis) in the other half. As was also seen in our collected data, the incidence of spontaneous resolution of extravasated fluid is substantially higher (70–80%) in patients without pancreatic necrosis.
The overall incidence of acute and chronic complications in our series of 169 acute attacks of pancreatitis is 5.3%. As expected, complications did not occur in the mild forms of grades A, B, and C pancreatitis, but occurred exclusively in the more severe forms, grades D and E pancreatitis, after the extravasation of pancreatic secretions. Even in this subgroup of 73 patients with fluid collections, the morbidity rate was relatively low in the absence of pancreatic necrosis, with an incidence of complications of 12.3%. Acute life-threatening complications such as hemorrhage, infected collections, and duodenal perforation were seen in only five patients, representing 6.8% of the 73 patients with fluid collections or 3.0% of 169 cases overall.
When the initial peripancreatic fluid collections are not absorbed, they tend to organize and slowly evolve into fully encapsulated collections called "acute pseudocysts." This evolution heralds the beginning of a potentially more complex and uncertain clinical course. The development usually takes more than 4 weeks, but because the timing is somewhat variable, the diagnosis is established only when a sharply defined circumferential capsule is clearly detected. As opposed to chronic pseudocysts, acute pseudocysts have a thin friable capsule and an unstable natural history. They can diminish or grow in size, resolve, rupture, drain into the pancreatic duct, or fistulize into the gastrointestinal tract. Spontaneous resolution has been reported in 40% of acute pseudocysts known to be present for less than 6 weeks, whereas they tend to remain stable when older than 12 weeks [16]. Complications such as rupture, hemorrhage, or infection have been reported in 18–50% of cases [16–19]. A follow-up CT series of 75 patients with acute pseudocysts reported enlargement or complications requiring surgery in about half and resolution or stable size in asymptomatic individuals in the other half [20].
Follow-up CT examinations in our 73 grades D and E patients with fluid collections documented five patients with single pseudo-cysts, for an incidence of 6.8%, or 3.0% of the entire group of 169 cases. Hemorrhage from rupture of a pseudocyst occurred in two patients, necessitating surgical intervention. The prevalence of hemorrhagic pseudocysts, similar to the two cases in our series, varies in different reports from 2% to 31% of acute pseudocysts [17, 21]. After an acute attack of pancreatitis, hemorrhage is usually not associated with ruptured pseudoaneurysms, which tend to occur later after an acute episode [22]. Rather, in the acute phase, hemorrhage most often occurs secondary to capillary bleeding in the wall of the pseudocyst or in the retroperitoneum. Because the natural history, clinical significance, and surgical management are uncertain, a conservative noninterventional approach, particularly for asymptomatic pseudocysts smaller than 5 cm, has been accepted in clinical practice [20]. Surgical or interventional drainage procedures are reserved for complications (such as hemorrhage or infection) and for symptomatic enlarging pseudocysts diagnosed by follow-up imaging studies.
Three individuals in our series of 169 patients developed chronic complications (two cases of chronic pancreatitis and one case of groove pancreatitis), for an incidence of only 1.8%. Groove pancreatitis was diagnosed when there was focal inflammation exclusively or predominantly involving the head of the pancreas and associated fluid in the groove between the head of the pancreas and the second portion of the duodenum [23]. Although these are important long-lasting, irreversible, and clinically debilitating developments [23, 24], their true incidence rate is difficult to establish without close patient supervision and repeated long-term follow-up examinations.
Our retrospective survey of 169 patients with attacks of nonnecrotizing acute pancreatitis has several limitations that may affect the veracity of our results. Because of its retrospective nature, this is not a controlled study, and long-term follow-up examinations in some of our patients with unresolved fluid collections and acute pseudocysts were not always available. Follow-up imaging studies were available for review in approximately 50% of our entire patient population and in 70% of patients with fluid collections, including 100% of grade E patients with larger collections and more severe and protracted clinical presentations. We likely underestimated the true incidence of acute and chronic complications because longer-term follow-ups in some of our patients with unresolved fluid collections and acute pseudocysts may have yielded additional complications. On the other hand, long-term follow-up examinations could have missed other unrecorded subliminal intervening acute episodes of pancreatitis, particularly in alcoholic patients, that might have contributed to the development of late complications. Patients may also have sought follow-up care and undergone imaging at an outside institution, and they may have developed complications of which we were not aware. In addition, despite the improved accuracy of MDCT examinations, small superficial patchy areas of pancreatic necrosis that might have contributed to the severity of the acute attack in the development of subsequent complications could have been missed.
In conclusion, in this series of 169 patients with milder forms of acute pancreatitis, 16 abdominal complications developed in nine patients, for an incidence of 5.3%. All complications occurred in grades D and E patients with fluid collections. Fluid was rapidly absorbed in most patients but led to complications in 12.3% in this subgroup of patients. Acute, life-threatening complications (hemorrhage, infection, perforation) were seen in five patients, for an incidence of 6.8%, among 73 episodes with fluid collections, or 3.0% of 169 total cases. Single acute pseudocysts were seen in five patients, and long-term chronic complications were documented in three patients. Severe abdominal morbidity can occasionally occur after an episode of acute pancreatitis in the absence of necrosis. In patients with milder interstitialforms of pancreatitis, routine follow-up CT examinations are indicated only in patients with fluid collections to document resolution of fluid or the development of complications.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  O. R. Brook, D. Fischer, D. Militianu, A. Eran, L. Guralnik, S. Israelit, and A. Engel Pseudopancreatitis in Trauma Patients Am. J. Roentgenol., September 1, 2009; 193(3): W193 - W196. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
