clinicalpathology of acute n-ecrotising pancreatitis · 72 appearance of acute necrotising...

J Clin Pathol 1986;39:68-74

Clinical pathology of acute n-ecrotising pancreatitisI NORDBACK,* K LAUSLAHTIt

From the Department of*Surgery and tPathology, University Central Hospital of Tampere, Finland

SUMMARY Seventy nine pancreatic specimens were obtained from patients treated with pancreaticresection for acute necrotising pancreatitis. The necrotising process had started in the periphery ofthe gland, so that eight of seventy nine cases contained peripancreatic (mainly fat) necrosis onlywithout any parenchymal necrosis. Peripheral parenchymal necrosis was characterised by a severe

inflammatory reaction, with multinucleated leucocytes and microabscess. In the deep parts of thepancreas coagulation necrosis was found. Vascular changes (thrombosis, vessel necrosis) correlatedwith postoperative haemorrhagic complications, but they did not seem to have had any importantrole in the necrotising process. The vascular changes seemed to be a secondary phenomenon.In clinical practice the most important aspects in reporting the histology of acute necrotisingpancreatitis are the extent of parenchymal necrosis, because the surgeon may overestimate itsextent, and the existence of vascular changes, because of the correlation with postoperativerecovery.

In recent years pancreatic resection has become afairly widespread treatment for acute necrotising pan-creatitis, expecially in Europe.`6 The precise indi-cations and contraindications for the procedure are,however, far from clear. The aim of pancreaticresection is to reduce the amount of necrotic tissue, tocut down the release of toxic substances, and toreduce the likelihood of infection. Whether the inci-dence of these sequences of pancreatitis is mosteffectively reduced is not yet known. For this reason

and because severe forms of pancreatitis with necrosisare quite rare most of the workers concerned with thiskind of treatment, including pathologists, have as yetvery limited experience of the procedure. We had an

opportunity to study 79 surgical pancreatic samplesobtained by resection from patients with acute necro-tising pancreatitis. The correlations between histol-ogy and certain clinical variables were investigated inthe first 50 patients from whom clinical data wereobtainable.

Material and methods

Seventy nine patients treated between 1973 and 1983who had undergone pancreatic resection for acutenecrotising pancreatis were studied. The age rangewas 23-75 years (mean 46 years). The aetiology of thepancreatitis was alcohol in 55 cases (70%), gallstonesin 13 cases (16%), and blunt abdominal trauma in

Accepted for publication 29 August 1985

one case. The aetiology of the remaining cases wasunknown.Laparotomy was performed in patients with pan-

creatitis who deteriorated despite intensive medicaltreatment and who had signs of peritonitis and inpatients with an uncertain diagnosis who were sus-pected of having visceral perforation. The criterionfor pancreatic resection was clear evidence of necrosisof the pancreas based on the macroscopical evalu-ation made by the surgeon during laparotomy. A leftpancreatic resection, starting with splenectomy, wasperformed, extending to the level of the portal vein orthe middle of the head of the pancreas. One pan-creaticoduodenectomy was performed.The specimens were immediately fixed in formalin

or frozen in liquid nitrogen. Four to eight transversesections were taken at regular intervals throughoutthe pancreatic specimen, the number depending on itslength. This dissection was carried out after fixing informalin, or while the specimens were still frozen; inthe latter case the tissue slices were then thawed informalin at room temperature to ensure simultaneousfixing. Routine paraffin sections were processed andstained with hematoxylin and eosin and with Herovicistain for identification of connective tissue.7

Necrosis in the specimen was evaluated in each sliceas a percentage of the necrosis in the whole section inboth parenchymal tissue (acini and islets) and para-pancreatic tissue (fat and fibrous connective tissue,which was either peripancreatic or septal). The meanpercentage of the different slices was calculated foreach patient and placed on a semiquantitative scale of

68

on March 7, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.39.1.68 on 1 January 1986. Dow

nloaded from

http://jcp.bmj.com/

Clinical pathology of acute necrotising pancreatitis

.% -_ 0. 4

Fig. I Inflammatory reaction spreading throughinterlobular septumfrom peripancreas into pancreas.Peripancreaticfat is necrotic (top right). Acinar structuresseem swollen with interstitial accumulation offluid. x 500.

69

0-25%, 26-50%, 51-75%, and 76-100%.Histological assessment was carried out in all 79

cases. Data on the postoperative course and compli-cations were obtained on the first '50 patients, inwhom the clinicopathological correlations were alsoinvestigated. Twenty nine cases had to be excludedfrom the clinical study because the data had not beenreliably obtained at the time of investigation. Anal-ysis was carried out using the x2 test.

Results

On macroscopical examination black areas or a uni-form black covering could be seen on the surface ofthe specimen. After dissection it was found that innine cases (I I %) the whole specimen was dark brownor black, while in the others the black areas weremainly concentrated in the peripheral parts of thespecimen. The paler areas of the specimen seemed tobe swollen.On histological examination features of acute pan-

creatitis were evident in each sample. The interstitialinflammation process was characterised by accumu-

'p.,L '*4,

4

Fig. 2 Staging ofpathological changes according to depth. Peripancreaticfat is necrotic haemorrhaged (top right). Mostsuperficial layer ofpancreatic parenchyme is necrotic with inflammatory cell reaction (top middle). Next layer containsrecently necrotisedparenchyme with pycnotic nuclei and disarrangement ofstructure (top left, bottom right and middle),whereasfollowing layer shows oedematous acinar and ductal structures (bottom left). x 500.



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

70

Fig. 3 Microabscess with dense accumulation ofinflammatory cells in liquefactive necrosis. x 80.

lation of fluid and leucocytes, of which most werepolymorphonuclear. The inflammatory reaction wasmore severe in the interlobular spaces and septal tis-sue than within lobules, giving an impression ofinflammation starting from the peripheral parts of thegland and protruding through the septa into thedeeper parts (Fig. 1).

In each case cell necrosis was also found. The mostvulnerable areas seemed to be the peripancreaticadipose tissue, from where the necrosis spreadthrough the septa towards the pancreatic parenchyma(Fig. 1). In the parenchyma necrosis withinflammatory cells was first seen in the superficial lay-ers of the gland; deep to this were swollen and dis-rupted acini, also with an accumulation ofinflammatory cells; and deeper still were preservedbut oedematous acini (Fig. 2). In some cases micro-abscesses were found. These were areas a few milli-meters in diameter that contained a dense accumu-lation of inflammatory cells, exudate, and fibrinousmaterial (Fig. 3). Bacteria could not be detectedmicroscopically. Thus the presence of inflammatory

Fig. 4 Coagulation necrosis in middle ofpancreas withslightly increased amount ofconnective tissue. x 200.

Nordback, LauslahtiTable 1 Extent ofnecrosis in pancreatic parenchymal andparapancreatic (mainly adipose) tissue in 79 patients studied

Degree ofnecrosis (%) No ofpatients (%)

Parenchymal:0-25 38 (48)26-50 10 (13)51-75 11(14)76-100 20(25)

Parapancreatic:0-25 5 (6)26-50 9(11)51-75 19 (24)76-100 46 (58)

cells was especially typical of superficial necrosis. Theoccurrence of vascular changes was no prerequisitefor this type of necrosis. Coagulative necrosis wasfound in the cases in which the deeper layers of thepancreas were necrotic (Fig. 4). Acinar necrosis wasnot detected in 8 cases (10%), although para-pancreatic necrosis had occurred. Table 1 summarisesthe distribution of the extent of necrosis in casesstudied.The vascular changes were numerous. Throm-

bosing vasculitis with segmental necrosis of the vesselwall was a common finding; the affected vessels weresmall arteries, arterioles, and capillaries. In contrast,thrombosis without phlebitis was noticed in the veins.Haemorrhages were almost always found, the mostsevere being clearly associated with disruption of thevessel wall due to necrosis (Fig. 5). Table 2 shows theincidence of these histological changes. In some of thespecimens an increased amount of connective tissuewas seen; necrosis was also found in these specimens.

.5 S~~~q

t4st!s;us A A

*t<,, 5,J

Fig. 5 Necrotising vasculitis and vascular wall disruptionfollowed by severe haemorrhage and secondary accumulationofinflammatory cells. x 200.



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


Table 2 Histologicalfindings in pancreatic specimensfrom79 patients studied with acute necrotising pancreatitis

Histologicalfindings No ofpatients (%)

Parenchymal necrosis 71 (90)Parapancreatic necrosis 79 (100)Increased amount of connective tissue 18 (23)Microabscesses 5 (6)Intravasal thrombosis 68 (86)Vascular wall necrosis 38 (48)Haemorrhages 71 (90)

Calcium deposits, flattened epithelium, or dilatedducts were not found.No histological difference was found between the

various aetiological groups (gallstones, alcohol,unknown). Table 3 shows that the extent of extensiveparenchymal necrosis was proportional to the lengthof time from onset of symptoms until operation. Theseverity of vascular necrosis increased in the earlystages of pancreatitis but then decreased as the symp-toms persisted.

Nineteen of the first 50 patients died (38%). Threedied from cardiac arrest, one from liver necrosis, onefrom bone marrow depression, and one from renalfailure and pneumonitis of toxic origin. The maincause of death was sepsis (seven patients), whichstarted as an abscess in the region of the pancreas.Only one of the patients with microabscesses in thepancreas, however, developed a postoperativeabscess. Of the eight cases of postoperative abscess,three occurred in glands with under 50% parenchy-mal necrosis and five in glands with over 50% paren-chymal necrosis. Six patients died from postoperativehaemorrhage. Haemorrhage was the only histologicalvariable that cQrrelated significantly with mortality; p< 0.001 (Table 4). Altogether, 17 patients requiredfurther surgery for postoperative haemorrhage aris-ing in the operated area. Histological evidence ofnecrosis and haemorrhage seemed to predict a higher

71

Table 4 Mortality in relation to histological changes in 50patients treated with pancreatic resectionfor acutenecrotising pancreatitis

Mortality* (%)

Pancreatic parenchymal necrosis <50% 11/34 (32)>50% 8/16(50)

Parapancreatic necrosis <50% 3/12 (25)> 50% 16/38 (42)

Increased amount ofconnective tissue Yes 5/15 (33)

No 14/35 (40)Microabscesses Yes 3/5 (60)

No 16/45 (36)Vascular thrombosis Yes 15/43 (35)

No 4/7 (57)Vascular wall necrosis Yes 9/24 (38)

No 10/26 (38)Haemorrhages Yes 19/48(40) (p < 0001)

No 0/2y P 01

*Total mortality 19/50 (38%).

Table 5 Histological vascular changes and incidence ofreoperation requiredfor postoperative haemorrhage in 50patients treated with pancreatic resectionfor acutenecrotising pancreatitis

Patients requiring reoperationfor haemorrhage

Vascular thrombosis Yes 12/43 (28) (p < 002)No 5/7 (71) fVascular wall necrosis Yes 13/24 (54) } < )No 4/26(15) <Haemorrhages Yes 17/48 (35) 1p<001No 0/2 f(<001

risk of postoperative haemorrhage. On the otherhand, vascular thrombosis seemed to predict a

significantly lower risk of postoperative hae-morrhage; p < 0-02 (Table 5).

Discussion

Both the gross pathology and the microscopical

Table 3 Incidence ofhistological changes occurring between onset ofsymptoms and date ofoperation in 5 patients withacute necrotising pancreatitis

Duration ofsymptoms

< 4 days 4-7 days > 7 days

(n = 30) (n = 9) (n = 11)

Over 50% parenchymal necrosis 8 (27) 1 (11 )* 7 (64)*Over 50% parapancreatic necrosis 23 (77) 7 (78) 8 (73)Increased amount of connective tissue 10 (33) 1 (11) 4 (36)Microabscesses 2(7) 1 (11) 2(18)Intravasal thrombosis 26 (87) 8 (89) 9 (82)Vascular wall necrosis 11 (37)t 8 (89)t- - 5 (45)ttHaemorrhages 29 (97) 8 (89) 11 (100)

Figures in parentheses are numbers (%).*(p < 0001); t(p < 0-01); tt(p < 002).



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

72

appearance of acute necrotising pancreatitis havebeen described in detail over the past 50 years.8-10Most of the studies were, however, based on studieson animals or necropsy specimens. It is well knownthat rapid autolysis of the pancreas by the glandsdigestive enzymes destroys the normal histology ofthe organ soon after the patient's death. The tech-niques of immediate fixation or freezing of the pan-creas after interruption of its blood supply used in ourstudy have only been possible since the introductionof pancreatic resection for treating acute necrotisingpancreatitis. Since Watt's first report in 196311 thismethod has been increasingly used. Nevertheless,only a few reports have been published on the histol-ogy of the pancreatic specimen.

In most studies on animals the pathology of necro-tising pancreatitis is characterised by liquefactivenecrosis with abundant leucocyte infiltration, whereasin pancreatitis in man the necrosis is consideredrather as coagulative necrosis from pancreaticinfarction.I213 Phat et al studied 20 total pan-createctomy specimens from patients with acute nec-rotising pancreatitis and found both coagulativenecrosis and liquefactive necrosis with a severeinflammatory cell reaction.14 The histology of thenecrosis was essentially similar in our study. Coagu-lative necrosis was especially common in the innerparts of the pancreas, while in the superficial paren-chyma leucocyte infiltration was severe.The material obtained for this study differs from

that used in earlier studies in its aetiological distribu-tion. Usually pancreatitis from gallstones constitutesabout two thirds and pancreatitis induced by alcoholone quarter of the diseases."1 The distribution in ourstudy, however, represents typical Finnish material,comprising principally men with heavy drinkinghabits. 16The histology of the different aetiological types of

pancreatitis agreed with the findings of Phat et al.14Boutelier found necrosis to be caudal in pancreatitisof gallstone or unknown aetiology but diffuse orcephalic in the remaining cases. Diffuse or patchynecrosis was rare among our patients: we found onlyone case with areas of necrotic and non-necrotic tis-sue side by side throughout the specimen and consid-ered this to be a diffusely necrotised organ. This mighthave been, however, only the preliminary stage oftotal necrosis, because the non-necrotic areas hadclearly swollen acini.

Necrosis was seen predominantly in peripancreatictissue or in the peripheral parts of the pancreas itself,while the inner parts were, in most cases, not necrotic.This poses the surgeon a difficult problem-namely,how to make therapeutic decisions based on thesuperficial appearance of the organ, which do notseem to correlate with changes deep in the organ.

Nordback, Lauslahti

Overestimation of the extent of necrosis is not uncom-mon.1819 This also explains the high number of casesin this study with pancreatic parenchyma that werenot greatly necrotic. We considered the histologicalbiopsy technique suggested by Phat et al14 too haz-ardous because of the possibility of pancreatic fistulaein patients in whom no resection was performed. It isnot known whether fine needle aspiration cytologywould give useful information.

In studies that analyse necrotising pancreatitispathoanatomically and pathophysiologically themost interesting aspects are the early histologicalchanges, and the features which distinguish irre-versible from reversible damage. In an electron micro-scopical study Helin et al showed acinar dilatationwith accumulation of amorphous material in thelumen of acini, disappearance of microvilli, and swell-ing of the apical parts of the acinar cells.20 We alsofound considerable swelling and even vacuolisation ofacinar cells, especially near the necrotic areas. Thesechanges, together with the loss of adhesion of epi-thelial cells from their basement membranes, are non-specific and are to be found in cellular damage ofvarious origins.The fact that the amount of necrosis increased with

the interval from the onset of symptoms suggests acontinuation of the necrotising process. Retro-peritoneal extension of the necrosis has been docu-mented.21 It occurs in about one third of patients,despite pancreatic resection, causing total or partialnecrosis of the pancreas remnant.22 In two thirds ofcases, however, extension of the necrosis ceased; thelimiting factors involved are unknown.

It has been suggested that the vascular changes(vasculitis, thrombosis) are responsible for the necro-sis, which has hitherto been regarded as pancreaticinfarction.23 The peripheral or subcapsular onset ofparenchymal necrosis may support this vascularhypothesis. In contrast, in this series 27 of 31 (87%)of patients with necrosis of over 50% of parenchymaltissue had vascular thrombosis, and no difference wasfound when these were compared with cases withunder 50% parenchymal necrosis (41 of 48 (85%).Indeed, there was one case with multiple thrombosisbut without noticeable parenchymal necrosis. More-over, eight cases with peripancreatic but without pan-creatic parenchymal necrosis were found. Thus thetheory of necrosis as a consequence of vascularchanges seems questionable. Thrombosing vasculitismay equally well be a secondary phenomenon, as sug-gested by Takada et al.24 For experimental pan-creatitis we prefer the term "tryptic" or enzymaticlytic necrosis, which describes the role of lipolytic andproteolytic enzymes in the necrotising process.25 Wefound early and extensive fat necrosis, suggesting anincreased function of lipoactive enzymes (lipase,



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


phospholipase A2).26 The superficial parenchymalnecrosis with severe inflammatory reaction resembledthat of tryptic necrosis found in certain animal mod-els.1213 It has also been suggested that the coagulativenecrosis typical of pancreatitis in man found in thisstudy, especially in the inner parts of the pancreas,might be caused by the destruction of cell membraneinduced by phospholipase activity.'213 Although wedo not consider that vascular changes are the basis ofthe necrosis, systemic circulatory disturbances withhypotension induced by pancreatitis, together withintrapancreatic vascular changes may, however, beimportant cofactors in the extension of pancreaticnecrosis, because the pancreas, like the kidney, is vul-nerable to hypoxia.27The prevalence of vascular lesions (48% with vessel

wall necrosis, 86% with thrombosis) exceeded thatdescribed earlier: the corresponding figures in the nec-ropsy study caried out by Blenkinsopp28 (31 cases)were 39% and 42%, respectively, and those of thestudy by Phat et al"4 (surgical specimen from 20patients) 40% and 50%, respectively. Giant cellarteritis"4 was not found in our patients.

Contrary to widely held belief, Machado et alrecently described histologically verified acute necro-tising pancreatitis in 12 patients suffering fromchronic calcifying pancreatitis.29 We made a similarobservation in our study, although we prefer the termpancreatic fibrosis rather than chronic pancreatitis,because no clear acinar or ductal changes were found.One patient with symptoms resembling chronic pan-creatitis and who had had four acute episodes ofoedematous pancreatitis developed acute necrotisingpancreatitis with 26-50% parenchymal necrosis. Healso had an appropriately increased amount of con-nective tissue in the specimen. Moreover, in seventeenother previously symptom free patients an increasedamount of connective tissue was observed (total inci-dence 23%). An even higher incidence of pancreaticfibrosis (23 of 31, 74%) was observed in necropsymaterial from patients who died of acute necrotisingpancreatitis.28 A few more patients developed paren-chymal necrosis of more than 50% in the group ofpatients without (26 of 61, 43%) than with (five of 18,28%) pancreatic fibrosis. This might be taken tomean that a chronically inflamed pancreas is notcapable of producing such large amounts of toxicsubstances as a previously healthy organ.

Interestingly, the only histological feature of prog-nostic importance was the occurrence of hae-morrhages, but, as only two of the 50 patients whowere studied clinically had no haemorrhages thefinding can not have great discriminatory value. Theextent of necrosis was not an important variable:three of eight patients died (mortality 38%) withoutnoticeable parenchymal necrosis. Our material, how-

73

ever, comprised only the severe forms of pancreatitis,and thus no comparison could be made with milderforms of the disease. Moreover, the extent of necrosisin the surgical specimen may not closely correlatewith the extent of necrosis in the remaining part of thepancreas, which governs the subsequent clinicalcourse.

Microabscesses in the pancreatic specimens did notpredict postoperative abscess. The reason might lie inthe focal nature of these changes, as well as in theabsence of bacteria on microscopy. The correlationsbetween vascular changes and postoperative hae-morrhage seemed logical: vascular wall necrosisincreased the danger of postoperative haemorrhageand the trend to thrombosis reduced it.To conclude, in clinical practice the most important

points- to be specified on the pathology report of thehistology of acute necrotising pancreatitis are theextent of pancreatic necrosis, and the presence of vas-cular damage. The first would be instructive for thesurgeon, who may have overestimated the extent ofnecrosis because of the severity of superficial pan-creatic damage. The second correlates with the post-operative clinical course.

Referen

'Hollender LF, Gillet M, Kohler JJ. Die dringliche Pan-kreatektomie bei der akuten Pankreatitis. Langenbecks ArchChir 1971;328:314-27.

2Kiimmerle F, Neher M, Sch6nborn H, Mangold G. VorzeitigenOperation bei Akuten hamorrhagisch-nekrotisierender Pan-kreatitis. Dtsch Med Wochenschr 1975;100:2241-5.

'Frey HP, Meyer W, Miller G, Maurer W. Subtotale distale Pan-kreatektomie in der Behandlung akuter nekrotisierender Pan-kreatitiden: Indikation und Resultate. HeIv Chir Acta1978;45:693-8.

'Jonsel G, Boutelier P. Observations during treatment of acute nec-rotizing pancreatitis with surgical ablation. Surg Gynecol Obstet1979;148:385-6.

'Alexander J-H, Guerrieri MT. Role of total pancreatectomy in thetreatment of necrotizing pancreatitis. World J Surg1981;5:369-77.

6Kivilaakso E, Lempinen M, Makeliinen A, Nikki P, Schr6der T.Pancreatic resection for acute fulminant pancreatitis. A ran-domized controlled study. Ann Surg 1984;199:426-31.

'Herovici C. Polychrome stain for differentiating precollagen fromcollagen. Stain Technol 1963;38:204-6.

'Quick B. Acute pancreatitis. Aust NZ J Surg 1932;2:115-40.9Popper HL, Necheles H, Russel KC. Transitional of pancreatic

edema into pancretic necrosis. Surg Gynecol Obstet1948;87:79-82.

'"Thal AP, Perry JF Jr, Egner W. A clinical and morphologicalstudy of forty-two cases of fatal acute pancreatitis. Surg Gyne-col Obstet 1957;105:191-202.

Watts GT. Total pancreatectomy for fulminant pancreatitis. Lan-cet 1963;ii:384.

"2CreutZfelt W, Schmidt H. Aetiology and pathogenesis of pan-creatitis. Scand J Gastroenterol 1970;5:47-62.

13Anonymous. Mechanism of pancreatic autodigestion. [Editorial].N Engl J Med 1970;283:487-8.

"4 Phat VN, Guerrieri MT, Alexander JH, Camilleri JP. Early histo-logical changes in acute hemorrhagic necrotizing pancreatitis. A



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

74retrospective pathological study of 20 total pancreatectomyspecimens. Path Res Pract 1984;178:273-9.

13Banks PA. Pancreatitis. New York: Plenum, 1979.16 Kyosola K, Fock G. The aetiology of acute pancreatitis. Annales

Chirurgiae et Gynaecologiae Fenniae (Supplementwn)1976;65:3-6.

17Boutelier P. Les lesions macroscopiques dans les pancreatitesaigues n6crosantes: Modalite et difficultes d'exploration chirur-gicale. Essai de classification. Ann Chir 1972;26:243-8.

18Leger L, Chiche B, Louvel A. Pancreatic necrosis and acute pan-creatitis. World J Surg 1981;5:315-7.

19Nordback I, Pessi T, Auvinen 0, Autio V. Determination ofnecrosis in necrotizing pancreatitis. Br J Surg 1985;72:225-7.

20Helin H, Mero M, Markkula H, Helin M. Pancreatic acinar ultra-structure in human acute pancreatitis. Virchow Arch (PatholAnat) Histol 1980;387:259-70.

21 MacLean N. The role of surviving pancreas in late fatalities ofacute pancreatitis. Br J Surg 1977;64:345-6.

22Nordback I, Auvinen 0, Pessi T, Autio V. Complications afterpancreas resection for acute necrotizing pancreatitis. Acta ChirScand (in press).

23Schmitz-Moorman P. Comparative radiological and mor-phological study of the human pancreas. IV. Acute necrotizing

Nordback, Lauslahtipancreatitis in man. Path res Praod 1981;171:325-35.

24Takada Y, Appert HE, Howard JM. Vascular permeabilityinduced by pancreatic exudate formed during acute pan-creatitis. Surg Gynecol Obstet 1976;143:779-83.

25Becker V. Pathological anatomy and pathogenesis of acute pan-creatitis. World J Surg 1981;5:303-13.

26Zieve L, Vogel WC, Kelly WD. Species difference in pancreaticlipolytic and amylolytic enzymes. J Appi Physiol 1963;18:77-82.

"Warshaw AL, O'hara PJ. Suspectibility of the pancreas to ischemicinjury in shock. Ann Surg 1978;188:197-201.

28Blenkinsopp WK. The liver and pancreas in acute necrotizing pan-creatitis. J Clin Pathol 1978;31:791-3.

29 Machado MCC, da Cunha JEM, Bacchella T, de Barros Mott C,Duarte I, Bettarello A. Acute pancreatic necrosis in chronicalcoholic pancreatitis. Dig Dis Sci 1984;29:709-13.

Requests for reprints to: Dr I Nordback, Department ofSurgery, University Central Hospital of Tampere 33520Tampere, Finland.



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

clinicalpathology of acute n-ecrotising pancreatitis · 72 appearance of acute necrotising...

Documents