Significant elevations of serum lipase not caused by pancreatitis: a systematic review

Abstract

Background

Many authors advocate lipase as the preferred serological test for the diagnosis of pancreatitis and a cut-off level of three or more times the upper limit of normal (ULN) is often quoted. The literature contains no systematic review that explores alternative causes of a lipase level over three times as high as the ULN. Such a review was therefore the objective of this study.

Methods

The EMBASE and MEDLINE databases (1985 to August 2013) were searched for all eligible articles. Predetermined data were extracted and independently analysed by two reviewers.

Results

In total, data from 58 studies were included in the final analysis. The following causes other than pancreatitis of lipase levels exceeding three times the ULN were found: reduced clearance of lipase caused by renal impairment or macrolipase formation; other hepatobiliary, gastroduodenal, intestinal and neoplastic causes; critical illness, including neurosurgical pathology; alternative pancreatic diagnoses, such as non-pathological pancreatic hyperenzymaemia, and miscellaneous causes such as diabetes, drugs and infections.

Conclusions

A series of differential diagnoses for significant serum lipase elevations (i.e. exceeding three times the ULN) has been provided by this study. Clinicians should utilize this knowledge in the interpretation and management of patients who have lipase levels over three times as high as the ULN, remaining vigilant for an alternative diagnosis to pancreatitis. The medical officer should be aware of the possibility of incorrect diagnosis in the asymptomatic patient.

Introduction

Pancreatitis is an inflammatory disorder of the pancreas, the clinical course of which ranges from mild disease to multiple organ dysfunction and death. Guidelines for the diagnosis of pancreatitis commonly stipulate the presence of two of the three following criteria: typical abdominal pain; characteristic computed tomography (CT) findings, and/or amylase and/or lipase levels of three or more times the upper limit of normal (ULN).¹

The sensitivity and specificity of serum lipase for pancreatitis at a cut-off of three times the ULN range from 64% to 100% and from 99% to 100%, respectively.^2–5 In comparison, at the same cut-off, the sensitivity and specificity of amylase range from 50% to 78.6% and from 99% to 100%.^2–5 Lipase provides increased sensitivity in the diagnosis of acute pancreatitis caused by alcohol consumption, and also in patients presenting later in their clinical course.⁶

Lipase is an enzyme that catalyses the breakdown of triglycerides.⁷ In addition to pancreatic acinar cells, lipase is found in the gastrointestinal tract, including the oesophagus, duodenum, stomach and colon.⁷ Lipase has also been described in the liver, heart, lungs and leukocytes.^7,8 Pancreatic lipase content is approximately 100 times that of the small intestine and liver, and the pancreas to serum lipase concentration gradient is close to 20 000.⁷

In addition to pancreatitis, mild lipase elevations above the ULN may be seen in a wide range of conditions, including peptic ulcer disease and other hepatobiliary disorders.⁶ However, to the authors' knowledge, no previous study has systematically evaluated the aetiology of lipase levels of three or more times the ULN.

Knowledge of the potential alternative causes of significantly raised lipase is essential for clinicians because such levels may be erroneously interpreted as indicative of pancreatitis in some cases. This is especially important as lipase is advocated as the preferred test for the diagnosis of pancreatitis in various guidelines and much weight is given to levels that exceed the ULN by three or more times.^1,9

The literature was therefore reviewed for causes other than pancreatitis of lipase levels of three or more times the ULN. The objective of this review is not to suggest an alternative ‘cut-off’ for raised lipase, but, rather, to alert the clinician to other potential causes of significant lipase elevation (i.e. levels of three or more times the ULN) in order to facilitate more accurate interpretation and improve patient care.

Materials and methods

Eligibility criteria

The criteria for the inclusion of a study in this systematic review required the reporting article to have been published after 1985. Lipase levels of three or more times the ULN were defined as representing significant elevation. At least one patient must have clearly shown lipase levels of three or more times the ULN attributable to a cause other than pancreatitis, except in exceptional circumstances as outlined below. Review articles, conference abstracts and proceedings, and articles which made insufficient explanation for the lipase result(s) were excluded from data analysis.

Literature search

Using the keywords ‘pancreatitis’, ‘lipase’, ‘amylase’ and ‘diagnosis’, the MEDLINE and EMBASE databases were searched for relevant articles published from 1985 to August 2013. Methods of lipase determination were initially labour-intensive and technically difficult, and more precise and faster assays were developed in the late 1980s; therefore, this study targeted articles published from 1985 onwards.^10–12 Additional searches were performed using MEDLINE and EMBASE based on manual checking of the reference lists from all appropriate articles. Any further studies were included if they met the inclusion criteria.

Study selection

Studies from the literature search were initially screened by reviewing article titles and abstracts. Studies were excluded if the review of the abstract identified any of the relevant exclusion criteria. When study inclusion based on article title and abstract was equivocal, the full text was reviewed and further screened for inclusion and exclusion criteria. Articles deemed suitable for inclusion were independently assessed by two reviewers before a final decision on study eligibility was made.

Data collection

A standardized table was utilized in the process of extracting data from each article. The information collected from each study included: author(s); study year; city(s)/country(s) of study; number of patients in the series (including numbers with elevated lipase, and numbers with lipase levels of three or more times the ULN); normal ranges of lipase and amylase; mean peak lipase levels; mean peak amylase levels; evidence of pancreatitis in CT and/or ultrasonography (US); presence of abdominal pain or tenderness; method(s) used to confirm diagnosis, and condition(s) associated with lipase levels of three or more times the ULN as cited in the article (Appendix 1). Obvious confounders, such as alcoholism or renal failure, were noted. Individual patients in whom these confounders coexisted with the cause under investigation were generally excluded, except in studies outlining renal failure or alcoholism as potential causes of significant lipase elevation.

Causes of lipase of three or more times the ULN in each study were categorized as either (i) ‘probable’ or (ii) ‘possible’. These two groups were defined according to: (i) the number of study(s) supporting the cause of raised lipase; (ii) the strength of confirmatory evidence, such as cross-sectional imaging, laparotomy/surgical findings and/or specific serological or other tests; (iii) the possibility of underlying subclinical pancreatitis based on the mechanism in question, and/or (iv) the presence or absence of other potential causes of raised lipase.

Formulation of results

Qualitative items from each study, primarily representing different causes of lipase levels of three or more times the ULN, were synthesized and divided into four final groups: (i) reduced clearance or physiological causes; (ii) intra-abdominal pancreatic and non-pancreatic causes; (iii) critical illness, and (iv) other. Neither a quantitative analysis nor a formal meta-analysis could be conducted as a result of the significant heterogeneity among the articles and because of the inclusion of case reports in the final analysis. This systematic review was completed in line with PRISMA (preferred reporting items for systematic reviews and meta-analyses) guidelines.¹³

Results

The process of study selection and exclusion is outlined in Fig. 1. Data from 58 studies were included in the final qualitative analysis from which probable causes other than pancreatitis for lipase levels of three or more times the ULN were drawn.

Figure 1.

Study selection process. ULN, upper limit of normal

Three of the 58 articles indicated potential associations with lipase levels of three or more times or close to three times the ULN without conclusive evidence (Table 1). Table 2 summarizes all potential causes of lipase levels of three or more times the ULN without associated pancreatitis, categorized into the four groups outlined above. The strength of evidence for each alternative cause is also outlined in Table 2.

Table 1.

Possible other causes of lipase levels of three or more times the upper limit of normal (ULN)

Authors, year	Possible association(s) with lipase of three or more times the ULN
Bokemeyer, 200240	IBD: three of 70 and three of 66 patients with UC and CD, respectively, had lipase levels over twice the ULN; however, the potential number of patients with lipase levels of three or more times the ULN is not explicitly recorded in the article
Carroccio., 200641	Coeliac disease: adults and children with coeliac disease had lipase elevations of, respectively, up to 2.5 and 2.1 times the ULN at disease diagnosis
Heikius., 199942	IBD: serum lipase was elevated by up to 2.9 times the ULN

CD, Crohn's disease; IBD, inflammatory bowel disease; UC, ulcerative colitis.

Table 2.

Summary of causes of significantly raised lipase without pancreatitis

Categories of causes	Strength of evidence: study type(s) and method(s) of diagnosis confirmation
Reduced clearance/physiological
Renal impairment	AKI: case series (×2), case–control study (×2); blood tests ± US, CT, laparotomy
AKI8,14-16	CRF: case series (×1), case–control study (×1), cohort study (×1); blood tests ± US +
CRF with or without haemodialysis or continuous ambulatory peritoneal dialysis11,49,50	?others (NR)
Macrolipase51-56	Macrolipase: case report (×6); blood tests ± US, CT, laparotomy, ERCP, urine amylase,
Potential complexes: IgG, IgA, IgM, α-2-microglobulin	urine lipase, chromatography, immunoprecipitation assays
Potential associated conditions: coeliac disease, Crohn's disease, hypergammaglobulinaemia, liver cirrhosis, multiple myeloma, systemic lupus erythematosus
Intra-abdominal pancreatic and non-pancreatic (non-traumatic)
Hepatobiliary	Biliary atresia: case series; blood tests + US ± MRI
Biliary atresia57	Cholecystitis: case series (×2), case–control study (×2); blood tests ± US, CT,
Cholecystitis11,14,15,43	laparotomy
Cholangitis15	Cholangitis: case–control study; blood tests ± US, CT, ?others (NR)
Liver necrosis8	Liver necrosis: case–control study; blood tests ± US, CT, laparotomy
?Post-ERCP24	Post-ERCP: case series; blood tests
Gastroduodenal	Gastric perforation: case report; blood tests + CT, laparotomy
Gastric perforation58	Peptic ulcer disease: case series (×2), case–control study (×1), cohort study (×2); blood
Peptic ulcer disease with or without perforation2,4,5,14,43	tests ± US, CT, MRCP, ERCP, endoscopy, laparotomy
Bowel	Bowel necrosis/perforation: case–control study (×2); blood tests ± US, CT, laparotomy
Bowel necrosis and/or perforation8,15	Bowel obstruction: case report (×1), case–control study (×1), cohort study (×1); blood
Bowel obstruction2,15,59	tests ± US, CT, MRCP, ERCP, laparotomy
Neoplasms	Acinar cell carcinoma: case report (×1), case series (×2); blood tests ± US, CT, biopsy,
Acinar cell carcinoma of pancreas60-62	laparotomy
Others: ?hepatocellular carcinoma, ?metastatic bowel cancer, ?metastatic gastric	Other neoplasms: case series with <10 patients (×1), case series (×1), case–control
cancer, ?others11,14,17	study (×1); blood tests ± US, CT, laparotomy
Non-neoplastic, pancreatic	Non-pathological pancreatic hyperenzymaemia:– observational study/case series (×4),
Non-pathological pancreatic hyperenzymaemia18-22	cohort study (×1); blood tests ± US, CT, MRI (including MRCP and MRCP-S), ERCP,secretin-cerulein test, urine amylase
Pancreas transplant rejection63	Pancreas transplant rejection: case series; blood tests + pancreatic biopsy
Other	Ruptured AAA: cohort study; blood tests ± US, CT, MRCP, ERCP
Ruptured AAA2	Haemorrhage from other sources: case report (×2); blood tests ± US, CT, laparotomy
Intra-abdominal haemorrhage from other sources – ruptured ectopic pregnancy, ruptured ovarian cyst, ?others64,65	Peritonitis: case–control study; blood tests + ?others (NR)
Peritonitis11
Critical illness
Neurosurgical	ICH: observational study/case series (×2), case–control study (×3); blood tests ± US, CT,
ICH with or without craniotomy8,25,66-68	craniotomy
TBI/head injury without intracranial haemorrhage25,67,69	TBI:– observational study/case series (×1), case–control study (×1), cohort study (×1); blood tests ± US, CT, craniotomy
?Other critically ill/intensive care unit patients, including those with multi-system trauma 25-27,67,68	See text and Appendix 1
Other
Diabetes	Type 1 DM: case series (×1), cohort study (×1); blood tests ± ?others (NR)
Type 1 DM (insulin-dependent diabetes mellitus)14,70	DKA: case series with <10 patients (×1), case series (×1), cohort study (×3); blood tests
DKA (adults and children)29-33	± CT, other imaging (not specified)
?Type 2 DM28	Type 2 DM: case–control study; blood tests + ?others (NR)
Drugs	Chronic alcoholism: case–control study; blood tests + ?others (NR)
Alcohol (chronic alcoholism)34	Morphine + prostigmine provocation: cohort study; blood tests + CT, ERCP, HIDA scan
Morphine + prostigmine provocation (in healthy controls and sphincter of Oddi	Sorafenib: case series; method of diagnosis confirmation NR
dysfunction)36	Nilotinib: case series; blood tests + ?others (NR)
?Sorafenib71	DPP-4 inhibitors: cohort study; blood tests only
?Nilotinib72
?DPP-4 inhibitors (sitagliptin and saxagliptin)35
?Others
Infection	HCV: case–control study; blood tests ± US
HCV37	HIV: observational study; blood tests ± US, CT
? HIV38	Gastroenteritis: case–control study (×2); blood tests ± US, CT, ?others (NR)
?Gastroenteritis11,15	CMV: case–control study; blood tests + ?others (NR)
?Others (e.g. ? CMV)11
Miscellaneous	Sarcoidosis: case report; blood tests + US, CT, ERCP, gel filtration chromatography, liver
?Sarcoidosis39	biopsy
?Renal transplant rejection11	Renal transplant rejection: case–control study; blood tests + ?others (NR)

^aSee text.

Possible causes of significantly raised lipase, as defined in the Methods section, are denoted by a preceding ‘?’.

‘Blood tests’ as one method of confirming a diagnosis implies the use of standard accepted serological tests for each respective condition (if applicable). Clinical assessment is by implication a part of diagnosis confirmation, and is thus not explicitly mentioned.

AAA, abdominal aortic aneurysm; AKI, acute kidney injury; CMV, cytomegalovirus; CRF, chronic renal failure; CT, computed tomography; DKA, diabetic ketoacidosis; DM, diabetes mellitus; DPP-4, dipeptidyl peptidase-4; ERCP, endoscopic retrograde cholangiopancreatography; HCV, hepatitis C virus; HIDA, hepatobiliary iminodiacetic acid; HIV, human immunodeficiency virus; ICH, intracranial haemorrhage; Ig, immunoglobulin; MRCP, magnetic resonance cholangiopancreatography; MRCP-S, secretin-enhanced MRCP; MRI, magnetic resonance imaging; NR, not recorded; TBI, traumatic brain injury; US, ultrasonography.

Alternative causes of lipase levels of three or more times the ULN

Reduced lipase clearance/physiological causes

There is no correlation between the cause of acute kidney injury (AKI) and raised lipase.^14–16

Intra-abdominal non-pancreatic and pancreatic causes (non-traumatic)

Hepatocellular carcinoma, bowel cancer with liver metastases and unspecified metastatic bowel and gastric cancer can all be considered as possible causes of lipase levels of three or more times the ULN.^11,14,17 Patients with these diagnoses in the respective series often had a normal pancreas on US and/or CT scanning, and a negative laparotomy when performed (Appendix 1).

Non-pathological pancreatic hyperenzymaemia is a more recently described phenomenon with good evidence as a cause of significant lipase elevation.^18–22 It is characterized by the absence of identifiable pancreatic disease in the presence of elevated pancreatic enzyme levels.²³

Elevated lipase levels may also be seen post-endoscopic retrograde cholangiopancreatography (ERCP), without any associated evidence of pancreatitis.²⁴ Gottlieb. showed that lipase levels of 1000 units/l (normal range: 40–240 units/l) after ERCP had specificity of only 55% for pancreatitis.²⁴ The existence of foci of pancreatitis cannot be disproved, however, and the clinical significance of this lipase rise is uncertain.

Lipase in critical illness

Multi-trauma patients without head injury have been reported to have significant lipase elevations without strong radiological or operative evidence for pancreatitis.^8,11,25–27 Broadly, these patients have sustained blunt abdominal or pelvic trauma, in some cases with associated liver injury.^8,11,25,27 However, it is very difficult to disprove underlying subclinical pancreatitis in such patients, and therefore multi-trauma cannot be thought of as a probable cause of lipase levels of three or more times the ULN without pancreatitis.

Some authors have also pointed to lipase elevations of three or more times the ULN in non-surgical intensive care unit (ICU) patients with such diagnostic categories as septic shock and respiratory failure; underlying pancreatitis remains a significant possibility here (Table 3).^25,26

Table 3.

Possible pathogeneses of alternative causes of significantly raised lipase

Reduced clearance/physiological

Renal impairment Lipase handling and metabolism: glomerular filtration is primarily responsible for the removal of lipase from the serum; the majority of lipase is then reabsorbed by renal tubules with subsequent intra-renal degradation.7,43,73 Liver, lungs and spleen also likely involved in lipase metabolism7,74 Raised lipase in renal impairment: reduced glomerular filtration at least partially explains this.50 Other mechanisms also seem to exist as lipase levels in individuals do not correlate well with changing renal function75 Haemodialysis and CRF patients: additional possibilities include subclinical foci of pancreatitis and increased lipase reabsorption in the intestine49,50

Macrolipase Enlarged lipase molecule: may artificially elevate serum lipase measurements caused by reduced renal filtration53,54 Presence of macrolipase may sometimes indicate an increased risk for associated disorders, such as autoimmune or lymphoproliferative disease51,52

Intra-abdominal pancreatic and non-pancreatic causes (other than pancreatitis)

Inflammation: pancreas probably highly sensitive to inflammation in adjacent organs/structures, such as the biliary tract and gastrointestinal tract, with subsequent increases in serum lipase caused by inflammation of these organs11,43 Obstruction: biliary regurgitation into the pancreas from distal obstruction may induce pancreatic inflammation; biliary, pancreatic duct or bowel obstruction may also cause increased bloodstream diffusion of pancreatic enzymes4,11,14 Alternative sources of lipase: stomach, small bowel, liver, gallbladder and other surrounding organs may potentially act as non-pancreatic sources of lipolytic enzymes14 Peritonitis: mechanisms not clear with regards to peritoneal irritation (e.g. intra-abdominal haemorrhage) and elevated serum lipase65 Impaired hepatic function: reduced hepatic metabolism of lipase may occur with impaired hepatic function (e.g. chronic liver diseases like biliary atresia)57,74 Neoplasia: lipase elevation may be related to biliary or pancreatic ductal obstruction, formation of macrolipases, or (functional) tumour mass lipase production/hypersecretion17,60,62 Non-pathological pancreatic hyperenzymaemia: mechanisms not yet characterized, although a genetic basis should be entertained23

Critical illness

ICU patients with critical illness/multi-organ failure: pancreatic hypoperfusion with cellular stress may be responsible for lipase elevations;25 alternatively, pancreatic enzymes in the gut may enter the submucosa and subsequently the systemic circulation during times of gut ischaemia from reduced splanchnic flow25,26 Neurosurgical patients with severe head injuries: exocrine pancreatic enzyme production may be activated through central pathways, such as increased vagal tone, altered adrenergic stimulation, and/or release of activating hormones such as cholecystokinin66,68

Some alternative causes

Diabetic ketoacidosis and type I DM Possible causes of raised lipase in DKA: may include relative renal hypoperfusion and thus reduced serum clearance of the enzyme;30 alternatively, impaired pancreatic and/or splanchnic flow, in addition to non-pancreatic release of lipolytic enzymes, may be contributing factors.17,54,56 In DKA and recent-onset type 1 DM in general: inflammation from immune-mediated β-cell destruction could potentially ‘spill over’ into the exocrine pancreas, with subsequent acinar cell damage14,30,70

CRF, chronic renal failure; DKA, diabetic ketoacidosis; DM, diabetes mellitus; ICU, intensive care unit.

Other causes of lipase levels of three or more times the ULN

Diabetes

Malloy. showed that a very small proportion of patients with asymptomatic type 2 diabetes had lipase levels of three or more times the ULN; this may thus be a possible cause of significant lipase elevation.²⁸

Abdominal pain is a common finding in diabetic ketoacidosis (DKA); however, in the studies outlined in Table 2,^29–33 DKA patients with elevated lipase and abdominal pain rarely showed clinical and/or radiological pancreatitis (see also Appendix 1).

Drugs

Despite the association of chronic alcoholism with chronic pancreatitis, some authors have shown lipase levels of three or more times the ULN in otherwise asymptomatic alcoholics.³⁴ Some patients also experienced significant abnormalities in serum lipase after the administration of the dipeptidyl peptidase-4 (DPP-4) inhibitors, sitagliptin and saxagliptin, for the treatment of type 2 diabetes.³⁵ None of these patients had clinical features of pancreatitis and lipase levels normalized in most after drug withdrawal.³⁵ Lobo. used intramuscular morphine (10 mg) and prostigmine (1 mg) for the investigation of sphincter of Oddi dysfunction (SOD).³⁶ Surprisingly, both patients with proven SOD and healthy controls showed a lipase rise that sometimes exceeded three times the ULN.³⁶ Provocation did not induce pain in any of the healthy controls.³⁶

Infection

Yoffe. found that some asymptomatic patients with chronic hepatitis C virus (HCV) infection demonstrated lipase levels of three or more times the ULN; lipase levels often normalized in patients who responded to HCV therapy.³⁷ Patients with human immunodeficiency virus (HIV) may also sometimes develop high lipase levels, although the contribution of the virus alone to this abnormality is debatable and at best possible.³⁸ Gastroenteritis is also a possible consideration in patients with elevated lipase without pancreatitis, but more evidence is required.^11,15

Miscellaneous causes

Sarcoidosis has been reported in a single case as a possible cause of lipase levels of three or more times the ULN.³⁹ This patient had no clinical or imaging features of pancreatitis.³⁹ Coeliac disease and inflammatory bowel disease (IBD) were associated with lipase elevations that came close to but did not reach three or more times the ULN^40–42 (Table 1).

Discussion

Summary of evidence

This systematic, qualitative review has outlined a wide variety of causes of lipase levels of three or more times the ULN with the purpose of significantly aiding the clinician in his or her interpretation of this result. Although the specificity of lipase testing for pancreatitis at a cut-off of three or more times the ULN is >99%, it is essential for health care providers to understand other potential causes of this abnormality.^2–5

As Table 2 shows, there are many potential causes other than pancreatitis for the significant elevation of lipase levels. Reduced clearance of lipase from the circulation can occur with renal impairment or macrolipase formation. Lipase may also be elevated as a result of other intra-abdominal pathologies arising from the stomach, bowel and hepatobiliary tract, and from neoplastic disease. Diabetes, drugs and infection can also be responsible for lipase levels of three or more times the ULN.

Table 3 explains the possible pathogeneses of some of these causes of elevated serum lipase.

Data bias and study limitations

As a result of the heterogeneous nature of the studies included, and often insufficient and poorly comparable quantitative data, quantitative analyses were not performed here. Some of the proposed causes of lipase levels of three or more times the ULN are at best speculative, but every effort has been made to include only those causes for which there are plausible mechanism(s) and for which there exists evidence from multiple sources. An attempt was made in the data collection process to account for underlying conditions that may confound lipase values, such as alcoholism and the presence of renal impairment. Although the methods by which diagnoses were confirmed were recorded, there will inevitably be cases in which underlying mild pancreatic inflammation may have been missed in the respective studies; this is especially likely in retrospective series.¹⁵ Normal findings in CT, ERCP, US and even laparotomy may not always preclude histological pancreatitis.^43–45 Indeed, the CT scan may be normal in up to 25% of patients with clinical pancreatitis.^30,46

Lipase assays

It is important for clinicians to understand that assay interference and the detection of non-pancreatic lipases is possible; this in itself will not usually produce levels of three or more times the ULN, but can definitely alter lipase results.^7,47 Demanet., for example, showed large increases in serum lipase after the administration of i.v. heparin to patients; the Ektachem and Wako assays were used in this study.⁴⁸ When assay interference is suspected, clinicians should contact their clinical chemistry laboratory for clarification.

Implications and future perspectives

Lipase remains a good serological test for the diagnosis of pancreatitis. However, as shown by this study, levels of three or more times the ULN do not automatically confirm the diagnosis and many possible differential diagnoses exist. The diagnosis of pancreatitis based on clinical and serological features will sometimes generate false positives as many of the alternative diagnoses described in the current review also present with abdominal pain.

In less clear circumstances, the simultaneous measurement of both amylase and lipase should also be considered, although many authors doubt that this practice enhances diagnostic accuracy.^1,2,6,15 Furthermore, the conditions presented here as causing lipase levels of three or more times the ULN often also induce elevations of serum amylase, which may negate any added diagnostic utility afforded by amylase measurement.⁶ Alternative serological tests may refer to amylase isoenzymes, immunoreactive trpysinogen (IRT) and elastase-1.⁶ The latter two tests offer little help over lipase or amylase as routine diagnostic tests for pancreatitis, although they may be helpful in the event that a diagnosis of pancreatitis is doubtful.⁶ Another potential course of action in cases of diagnostic difficulty is to lower the threshold for abdominal CT, which may pick up alternative causes for the patient's presentation.

Overall, elevated lipase levels, like raised amylase levels, can occur in both asymptomatic patients and patients with a myriad of differing medical and surgical conditions. Knowledge of these different aetiologies will aid clinicians in the decision making and treatment of patients with significantly raised lipase, especially in patients with equivocal lipase levels.

Conflicts of interest

None declared.

Study data

Author(s), year	Hospital, country Type of study	Patients in series and patients with raised lipase levels, n	Normal range of lipase	Normal range of amylase	Peak amylase level (non-pancreatitis), mean (range)	Peak lipase level (non-pancreatitis), mean (range)	Imaging (CT ± others)	Presence of abdominal pain/tenderness
Akaza., 200771	Fukuoka, Osaka, Sapporo, Tsukuba and Tokyo, Japan Prospective non-randomized open label study	129 73 had ↑ lipase w/o pancreatitis 8 had levels >5 × ULN, and 32 had levels >2–5 × ULN	NR	NR	NR	NR	NR	0/73 patients with ↑ lipase
Argiris., 199938	New York, USA Prospective observational study	86 26 had ↑ lipase w/o pancreatitis Number with lipase levels ≥ 3 × ULN NR	10–70 U/l	25–125 U/l	117 (range 17–806)	56 (range 9–482)	No US or CT features of pancreatitis in 6 patients with ↑ lipase; 1 patient had mild pancreatic duct dilatation and ‘minimal pancreatic enlargement;’ imaging not performed in others due to no clinical suspicion	5/86 patients with ↑ lipase (epigastric discomfort or dyspepsia) (pancreatitis was not suspected)
Ashley & Lauwers, 200260	Boston, USA Case report	1 Had lipase level ≥ 3 × ULN	NR	NR	42	1614	No CT, US or MRI features of pancreatitis	Nil
Asvesta., 198834	Thessaloniki, Greece Case–control study	168 72 had ↑ lipase At least 10 had lipase levels ≥ 3 × ULN	0–160 U/l	NR	NR	366 (79 in controls)	NR	0/168 patients with ↑ lipase
Bokemeyer, 200240	Minden, Germany Prospective observational study	136 19 had ↑ lipase w/o pancreatitis (11 had CD, 8 had UC) 6 had lipase levels >2 × ULN w/o pancreatitis	0–60 U/l	28–100 U/l	NR (1 patient had levels > 2 × ULN)	NR	No US features of pancreatitis in all 19 patients with ↑ lipase	0/19 patients with ↑ lipase
Brooks., 200959	Cleveland, USA Case report	1 Had lipase level ≥ 3 × ULN	12–70 U/l	0–137 U/l	1241	2664	No CT features of pancreatitis	Yes (epigastric pain)
Carroccio., 200641	Palermo, Italy Prospective observational study	202 (90 adults, 112 children) 19 adults and 21 children had ↑ lipase w/o pancreatitis No patient had lipase levels ≥ 3 × ULN (peak was 2.5 × ULN)	20–270 IU/l (adults); variable for children	10–160 IU/l (adults); variable for children	240 (median) (adults); 1.4 × ULN (median) (children)	432 (median) (adults); 1.3 × ULN (median) (children)	No US features of pancreatitis in all 202 patients	10/90 adults and 9/112 children (‘recurrent abdominal pain’)
Catton & Lobo, 201058	Nottingham, UK Case report	1 Had lipase level ≥ 3 × ULN	23–300 IU/l	30–110 IU/l	380	4398	No CT features of pancreatitis	Yes (epigastric pain)
Chase., 199643	Chattanooga, USA Retrospective case series	306 208 had non-pancreatic abdominal pain 26/208 had ↑ lipase 4 had lipase levels ≥ 3 × ULN	5–208 U/l	30–110 U/l	N/Ad(maximum 385)	N/Ad(maximum 3685)	NR	26/26 patients with ↑ lipase
Chen., 199649	Taipei, Taiwan Retrospective case series	84 22 had ↑ lipase w/o pancreatitis 4 had levels ≥ 3 × ULN w/o pancreatitis	<190 U/l	<180 IU/l	127 ± 17 (haemodialysis patients); 192.1 ± 24.5 (no dialysis)	206 ± 35 (haemodialysis patients); 146.1 ± 18.4 (no dialysis)	No US features of pancreatitis in all 22 patients with ↑ lipase	0/22 patients with ↑ lipase
Denz., 200725	Berlin, Mannheim and Karlsruhe, Germany Prospective observational study	66 52 had ↑ lipase 38 had lipase levels ≥ 3 × ULN	<150 U/l	NR	NR	NR	CT showed features of pancreatitis in 7/20 patients in which it was done; No CT features of pancreatitis in 13 patients with lipase ≥ 3 × ULNbb	NR
Diani., 199817	Pavia and Varese, Italy Retrospective case series	4 All 4 had ↑ lipase w/o pancreatitis 3 had lipase levels ≥ 3 × ULN w/o pancreatitis	8–57 U/l	NR	Within normal limits (all patients)	197 (patient 1); 120 (patient 2); 351 (patient 3); 412 (patient 4)	No US features negative of pancreatitis in all patients; no CT features of pancreatitis in at least 2 patients	NR
Duerksen., 200039	Winnipeg, Canada Case report	1 Had lipase level ≥ 3 × ULN	<190 IU/l	<120 IU/l	350	2000	No CT or US features of pancreatitis	Nil
Frank & Gottlieb, 199914	Landshut, Germany; Indianapolis, USA Retrospective case series	25 All 25 had ↑ lipase All 25 had lipase levels ≥ 3 × ULN	40–240 U/l	25–115 U/l	N/Ad	N/Ad	NR	19/25 patients with ↑ lipase
Gomez., 20122	Nottingham, UK Retrospective cohort study	2979 18 had lipase levels ≥ 3 × ULN w/o pancreatitisi	0–300 U/l	0–100 U/l	NR	NR	No CT features of pancreatitis in all 18 patients	18/18 patients (acute abdominal pain)
Goto., 200051	Akita and Matsumoto, Japan Case report	1 Had lipase level ≥ 3 × ULN	29–220 IU/l	115–360 IU/l	625	3338	No US features of pancreatitis	Nil
Gottlieb., 199624	Indianapolis, USA Prospective case series	231 Number with ↑ lipase NR 193 patients did not have pancreatitis	40–240 U/l	25–115 U/l	200 ± 13 (cf. 657 ± 104 for pancreatitis patients)	2444 ± 288 (cf. 9879 ± 1480 for pancreatitis patients)	NR	70/193 patients w/o pancreatitis had abdominal pain
Gullo, 199618	Bologna, Italy Prospective case series	18 All 18 had ↑ lipase At least 6 had lipase levels ≥ 3 × ULN	24–270 IU/l	0–220 IU/l	NR (range 1.4–4.1 × ULN)	NR (range 1.5–7.7 × ULN)	No CT or US (×2) features of pancreatitis in all patients	0/18 patients
Gullo, 200019	Bologna, Italy Prospective observational study	102j 19 had ↑ lipase w/o pancreatitis 7 had lipase levels ≥ 3 × ULN w/o pancreatitis	24–270 IU/l	0–220 IU/l	NR (range 1.3–5.2 × ULN)	NR (range 1.6–18 × ULN)	No US features of pancreatitis in all 19 patients with ↑ lipase	0/19 patients with ↑ lipase
Gullo., 200620	Bologna, Italy Prospective observational study	18w 16 had ↑ lipase w/o pancreatitis 7 had lipase levels ≥ 3 × ULN w/o pancreatitis	13–60 IU/l	13–53 IU/l (pancreatic isoamylase)	NR (range ∼0.8–3.9 × ULN) (pancreatic isoamylase)	NR (range ∼0.8–6.4 × ULN)	No US, CT or MRI features of pancreatitis in all patients	0/18 patients
Gullo, 200721	Bologna, Italy Prospective observational study	42w 38 had ↑ lipase w/o pancreatitis 30 had lipase levels ≥ 3 × ULN w/o pancreatitis at some point during study	13–60 IU/l	28–110 IU/l	NR	255 (mean of peak levels in each patient over 5 days)	No US or CT features of pancreatitis in all patients; No MRI features of pancreatitis in 36 patients	0/42 patients
Gumaste., 19934	New York, USA Prospective cohort study	170 (95 w/o pancreatitis + 75 with pancreatitis) 10 had ↑ lipase w/o pancreatitis 1 had lipase levels ≥ 3 × ULN w/o pancreatitis	31–213 U/l	?40–100 U/l	N/Ad	N/Ad	NRf	95/95 patients w/o pancreatitis (acute abdominal pain)
Haddad., 200429	Indianapolis, USA Prospective cohort study	50 20 had ↑ lipase w/o pancreatitis 7 had lipase levels ≥ 3 × ULN w/o pancreatitis	15–120 U/l (vitros); 5–51 U/l (colorimetric)	25–115 U/l (vitros); 25–115 U/l (Catachem)	162 ± 152 in DKA patients with ↑ lipase or amylase (cf. 38 ± 10 for DKA patients with normal enzymes)	404 ± 385 in DKA patients with ↑ lipase or amylase (cf. 46 ± 27 for DKA patients with normal enzymes)	CT features of pancreatitis in 1 patient (performed in 3 patients with lipase ≥ 3 × ULN)	36/50x patients (abdominal pain and/or vomiting)
Heikius., 199942	Oulu, Finland Prospective observational study	237aa 16 had ↑ lipase w/o pancreatitis No patient had lipase levels ≥ 3 × ULN (peak was 2.9 × ULN)	<160 U/l	70–300 U/l	176 (range 1.1–1.9 × ULN)	66 (range 1.1–2.9 × ULN)	US showed increased/patchy echogenicity of the pancreas in 6/38 patients with ↑ lipase or amylase	NR
Justice., 199466	Savannah, USA Prospective case series	38 (37 w/o pancreatitis + 1 with pancreatitis) 24 had ↑ lipase w/o pancreatitis Number with lipase levels ≥ 3 × ULN NR	23–208 U/l	30–110 U/l	140 ± 109	712 ± 614	No CT features of pancreatitis in the 3 patients it was performed in	NR
Keating & Lowe, 200252	St Louis, USA Case report	1 Had lipase level ≥ 3 × ULN	20–200 U/l	15–130 U/l	1143 (range 119–1143 over 2 months)	1139 (range 694–3095 over 2 months)	No CT or US (×2) features of pancreatitisg	Yes (lower abdominal cramps)
Klassen., 199663	Baltimore, USA Prospective case series	69 (68 w/o pancreatitis + 1 with pancreatitis) Number with ↑ lipase NR	23–208 IU/l	30–110 U/l	341 ± 40	1548 ± 222	NR	NR
Klimstra., 199261	New Haven and Washington, USA Retrospective case series	28 4 had ↑ lipase At least 1 had lipase levels ≥ 3 × ULN	0–1 Teitz units	NR	NR	NR (maximum 20 Teitz units)	NR	NR
Lando., 201235	Alexandria and Washington, USA Retrospective cohort study	90 31 had ↑ lipase w/o pancreatitis Number with lipase levels ≥ 3 × ULN NRh	∼60–65 U/l (ULN)h	NR	NR (maximum was 100 U/l above normal)	NR (maximum 453)	No imaging done (patients asymptomatic)	0/90 patients
Lee., 201067	Taipei, Taiwan Case–control study	89 43 had ↑ lipase or amylase (exact number with ↑ lipase NR)	≤200 U/l	≤190 U/l	NRt	NRt	See footnotey	Difficult to assess/NRv
Liu., 200168	Chicago, USA Case–control study	75 11 had ↑ lipase and amylase levels Number with lipase levels ≥ 3 × ULN NR	0–55 U/l	0–125 U/l	402	474	US done in 3/11 patients with ↑ lipase, US + CT done in 3/11 patients; nil imaging in 5 patients – no US or CT features of pancreatitis in these patients	0/11 patients with ↑ lipasek
Lobo., 200736	Nottingham, UK Prospective cohort study	44 (20 controls + 24 patients with SOD) At least 17 controls and 23 SOD patients had ↑ lipase levels Exact number with lipase levels ≥ 3 × ULN NR	23–300 U/l	30–110 U/l	∼7 × ULN (in controls, post-provocation with morphine and prostigmine)	∼10 × ULN (median – in controls, post-provocation with morphine and prostigmine); levels increased up to 50 × ULN in SOD patients (exact values NR)	NR	0/20 controls (post-provocation); 20/24 SOD patients (post-provocation)
Lott., 198611	Columbus, USA Case–control study	156 Up to 39 patients had lipase levels ≥ 3 × ULN w/o pancreatitise	Variablee	Variablee	N/Ad	N/Ad	NR	NR
Lott & Lu, 19918	Columbus, USA Case–control study	100 Exact number with ↑ lipase NR At least 6 patients had lipase ≥ 3 × ULN w/o pancreatitis	<200 U/l	<140 U/l	N/Ad	N/Ad	NR	NR
Malloy., 201228	San Diego, USA Case–control study	1869 119 type 2 DM patients and 29 obese non-DM patients had ↑ lipase w/o pancreatitis 13 type 2 DM patients and 2 obese non-DM patients had lipase levels ≥ 3 × ULN w/o pancreatitiss	0–60 U/l	13–53 IU/l (pancreatic amylase)	28 ± 20 (type 2 DM, screening), 27 ± 16 (type 2 DM, baseline); 26 ± 10 (obese, screening), 26 ± 10 (obese, baseline)	42 ± 29 (type 2 DM, screening), 43 ± 33 (type 2 DM, baseline); 33 ± 26 (obese, screening), 33 ± 14 (obese, baseline)	NR	0/119 patients with ↑ lipase
Manjuck., 200526	NY, USA Prospective cohort study	245 99 had ↑ lipase (11/99 had pancreatitis) Number with lipase levels ≥ 3 × ULN NR	23–208 U/l	NR	140 (cf. 389 for patients with image-proven pancreatitis)b	900 (cf. 2231 for patients with image-proven pancreatitis)b	US features of pancreatitis present in 2 patients; CT features of pancreatitis present in a further 9 patientsa	Difficult to assessc
Masoero., 199650	Torino, Italy Prospective cohort study	212 (63 with CRF, 98 on HD, 28 on CAPD, 23 with renal transplants) 118 had ↑ lipase levels 13 had lipase levels ≥ 3 × ULN (10 HD, 1 CAPD, 2 CRF)	23–300 U/l	0–200 U/l	390 ± 336 in HD patients; 280 ± 128 in CRF patients; 255 ± 109 in CAPD patients; 209 ± 93 in RT patients; 148 ± 64 in controls	389 ± 347 in HD patients; 292 ± 211 in CRF patients; 259 ± 177 in CAPD patients; 175 ± 77 in RT patients; 122 ± 62 in controls	US done in 40 patients with particularly high pancreatic enzyme levels – no US features of pancreatitis in these patients	0/212 patients
Matos., 200962	Indianapolis and Nashville, USA; Mannheim and Dresden, Germany Retrospective case series	17 4 had ↑ lipase levels w/o pancreatitis Number with lipase ≥ 3 × ULN NR	NR	NR	NR	NR (maximum 4151)	NR	10/17 had abdominal pain (unclear if same as those patients with ↑ lipase)
Mitura & Romanczuk, 200964	Siedlce, Poland Case report	1 Had lipase level ≥ 3 × ULN	13–60 IU/l	28–100 IU/l	2113	3184	No US features of pancreatitis (CT not done)	Yes (para-umbilical pain)
Nair., 200030	New York, USA Prospective case series	100 20 had ↑ lipase w/o pancreatitis 17 had lipase levels ≥ 3 × ULN; of these, 9 did not have pancreatitis	0–72 IU/l	0–330 IU/l	NR	NR	CT done in the 17 patients with lipase levels ≥ 3 × ULN – 8/17 had CT features of pancreatitis	27/100 patients (number with ↑ lipase and abdominal pain NR)
Nsien., 199231	Washington, USA Retrospective case series	3 All 3 had lipase levels ≥ 3 × ULN	23–208 IU/l	20–150 IU/l	497	∼3200	No CT features of pancreatitis in any patient	0/3 patientsl
Okumura., 199853	Otsu, Japan Case report	1 Had lipase level ≥ 3 × ULN	5–160 IU/l	25–80 IU/l	73	1730	No CT or US features of pancreatitis	Yes (NB lipase measured during Crohn's disease relapse)
Pezzilli., 199769	Bologna, Italy Prospective cohort study	88 6 had ↑ lipase 1 had lipase levels ≥ 3 × ULN	24–270 IU/l	0–220 IU/l	NR (range 22–728)	NR (range 5–826)	No US features of pancreatitis in all 6 patients with ↑ lipase	0/6 patients with ↑ lipase
Quiros., 200832	Davis, Aurora and Stanford, USA Prospective cohort study	67 21 had ↑ lipase w/o pancreatitis; 14 of these had abdo pain, and 7 did not) Number with lipase ≥ 3 × ULN NR	13–51 U/l	30–140 U/l	136 ± 295	94 ± 163	No imaging was done	42/67 patients (upon presentation); 14/21 patients with ↑ lipase
Rosti., 200972	Bergamo, Bologna, Brescia, Catania, Catanzaro, Naples, Rome, Turin and Udine, Italy Prospective case series	73 21 had ↑ lipase w/o pancreatitis 3 had lipase levels > 5 × ULN w/o pancreatitis	NR	NR	NR	NR	NR	0/21 patients with ↑ lipase
Ryan., 199427	Iowa, USA Prospective cohort study	100 17 had ↑ lipase (5 of these had clinical pancreatitis) Number with lipase ≥ 3 × ULN NR	<190 U/l	<170 U/l	544 ± 112 in patients w/o pancreatitis (but raised enzymes) cf. 725 ± 202 in patients with clinical pancreatitis	716 ± 150 in patients w/o pancreatitis (but raised enzymes) cf. 882 ± 349 in patients with clinical pancreatitis	No CT features of pancreatitis in the 17 patients with ↑ lipase; US features of pancreatitis in 1/17 patients with ↑ lipasey	5/17 patients with ↑ lipasez
Semakula., 199670	Bonheiden, Brussels, Ghent and Liege, Belgium; Stockholm, Sweden Prospective cohort study	307 Numbers with ↑ lipase NR)	0–9 years: 17–136 U/l; 10–19 years: 30–167 U/l; 20–39 years: 46–235 U/l	15–97 U/l	47 ± 21 (range NR)	88 (range 12–4,177)	NR	NR
Sinha., 201065	London and Southend-on-Sea, UK Case report	1 Had lipase level ≥ 3 × ULN	5–65 IU/l	25–125 IU/l	1121	1700	No CT features of pancreatitis	Yes (epigastric, then generalized)
Smith., 200515	Sydney, Australia Case–control study	1880 427 had ↑ lipase w/o pancreatitisr 62 had lipase levels ≥ 3 × ULN w/o pancreatitis	<190 U/l	27–100 U/l	<100 (median)	190–285 (median)	NR	NR
Sutton., 20095	Nottingham, UK Case–control study	1520 Number with ↑ lipase NR 41 had lipase levels ≥ 3 × ULN w/o pancreatitis	<300 IU/l	<110 IU/l	NR	NR	NR	NR
Taes., 200054	Ghent, Belgium; Columbus, USA Case report	1 Had lipase level ≥ 3 × ULN	23–300 U/l	30–110 U/l	∼300	3764	No CT features of pancreatitis (ERCP showed minimal dilation of choledochus and Wirsung duct)	Yes (patient initially admitted with chronic constipation and progressive abdominal distension with lipase of 575)
Testoni., 200922	Milan, Italy Prospective cohort study	25 11 had ↑ lipase Number with levels ≥ 3 × ULN NR	0–60 IU/l	13–53 IU/l (pancreatic amylase)	NR (range 76–227)	NR (range 65–364)	0/25 had US, CT or MRCP features of pancreatitis; 2/25 had MRCP features of chronic pancreatitis; 8/25 had MRCP-S features of chronic pancreatitis	0/25 patients
Vantyghem., 199933	Lille, France Prospective cohort study	164 (52 had DKA, 90 had poorly-controlled DM, 22 had well-controlled DM, 27 were controls) 31 had ↑ lipase w/o pancreatitis (19/52 patients with DKA had ↑ lipasem); Number with lipase levels ≥ 3 × ULN NR	0–190 IU/l	0–160 IU/l	194 ± 270 (in DKA group)	348 ± 690 (in DKA group)	NRn	NRo
Wen., 200557	Taipei, Taiwan Prospective case series	37 12 had ↑ lipase w/o pancreatitis 5 had levels ≥ 3 × ULN w/o pancreatitis	<120 U/l	<100 U/l	NR (although no child had levels ≥ 3 × ULN)	NR	0/37 had US features of pancreatitis; 18/31 patients with BA underwent MRI and did not have any features of pancreatitis	0/37 patientsq
Yoffe., 200337	Houston, USA Case–control study	103 26 had ↑ lipase w/o pancreatitis; Number with lipase levels ≥3 × ULN NR	114–286 U/l	25–115 U/l	82 ± 30 (cf. 75 ± 45 for controls)	253 ± 72 (cf. 210 ± 42 for controls)	NRp	0/103 patients
Yuki., 199955	Izumo, Japan Case report	1 Had lipase level ≥ 3 × ULN	0–50 IU/l	50–170 IU/l	Peak NR (but > ULN)	888	No US or CT features of pancreatitis	Nil
Zachee., 198516	Antwerp, Belgium Prospective case series	40 26 had ↑ lipase levels (3 of these patients had pancreatitis) Number with lipase levels ≥ 3 × ULN NR	NR	NR	4.6 ± 3.1 × ULN (w/o pancreatitis)	3.5 ± 3.9 × ULN (w/o pancreatitis)	NR	0/37 patients w/o pancreatitis
Zaman., 199456	Leuven, Belgium Case report	1 Had lipase level ≥ 3 × ULN	23–208 U/l	30–85 U/l	1104	2600	NR	NR

^aImaging performed in a total of 50 patients with elevated lipase levels.

^bMaximum mean amylase or lipase levels in the 39 of 50 patients with elevated lipase who underwent imaging but showed no radiologic features of pancreatitis.

^cThe majority of patients with elevated lipase levels were intubated and sedated, and thus clinical features (e.g. abdominal tenderness) were hard to ascertain.

^dMean for elevated lipase or amylase not provided as study explored non-uniform sets of patients with multiple different diagnostic categories (e.g. patients with abdominal pain compared with patients with pancreatitis).

^eExact number cannot be expressed as four different methods of lipase determination and two different methods of amylase measurement were utilized; not all four methods of lipase measurement were always used in every patient; three to six of 14 patients with cholelithiasis and/or cholecystitis, four to 13 of 43 patients with inflammatory disorders of the biliary and gastrointestinal tract, two to five of 35 patients with complications of renal transplantation, two to seven of 17 patients with renal failure, and two to eight of 22 patients with accidental trauma had lipase levels of ≥3 × ULN.

^fIt is not specified whether any of the 10 patients with raised lipase underwent CT scanning; the three patients with biliary pathology had a normal pancreas.

^gOne abdominal CT scan, and two US scans were performed over a 2-month period.

^hExact value not recorded.

ⁱTen patients had a ruptured AAA, six had a perforated duodenal ulcer and two had intestinal obstruction.

^jDifferent patient subset from Gullo, 1996.¹⁸

^kThe number of patients who were intubated/ventilated and/or had impaired sensorium (and thus would alter the reporting of abdominal pain or tenderness) was not reported in the article.

^lOne of these patients was confused and one was drowsy on admission, which may have made abdominal pain difficult to assess.

^mNine of 52 DKA patients had chronic alcoholic pancreatitis with pancreatic calcifications, but not all patients with very high lipase levels in the DKA group had chronic pancreatitis.

ⁿAlthough the number of patients who underwent CT scanning/other imaging was not recorded, the number of patients with ‘pancreatic calcifications’ is recorded in the text.

^oAlthough the number of patients with abdominal pain was not recorded, the article text does state that ‘none of the [DKA] patients had clinical signs of acute pancreatitis’; this statement is not further elaborated in the text.

^pAlthough not recorded, those patients with lipase levels >2 × ULN were assessed by US.

^qAbdominal pain may have been difficult to assess in some patients as the patients were all children, with an age range of 4 months to 11.2 years (median: 2.4 years).

^rA diagnosis of severe pancreatitis was made in case of US or CT changes of acute pancreatitis; a diagnosis of indeterminate pancreatitis was made in case of abdominal pain consistent with pancreatitis but no CT or US features of the disease.

^sThis study looked at two separate groups that had been recruited for other large clinical trials: type 2 DM patients, and obese non-DM patients; lipase and pancreatic amylase levels were measured at the study screening period, then at baseline prior to commencement of the other trials. Note that by the time of randomization (i.e. at baseline), only two (of 13) and non (of two) type 2 DM and obese non-DM patients, respectively, still had lipase levels ≥ 3 × ULN.

^tAlthough there were no specific recordings for amylase and lipase levels in patients without pancreatitis compared with patients with pancreatitis, some important results were as follows: mean peak amylase and lipase levels in severe head injury were 460 ± 544 and 1027 ± 1643, respectively; in spontaneous ICH were 362 ± 253 and 854 ± 1143, respectively; in SAH and other vascular disease were 337 ± 256 and 552 ± 853, respectively; and in patients who had tumour bleeding during admission/operation were 341 ± 381 and 749 ± 972, respectively. With respect to the type of neurosurgical intervention, the highest lipase levels were seen upon aneurysm clipping or other neurovascular surgery (551 ± 825) and upon insertion of external ventricular drains, intracranial pressure monitors and/or ventriculoperitoneal shunts (547 ± 680).

^uFour of 14 patients with elevated amylase and/or lipase had an abnormal pancreas on US, and seven of 11 had an abnormal pancreas on CT; in total, CT and/or US scanning were performed in 34 patients (four patients underwent both US and CT scanning; in total nine patients were radiologically positive for pancreatitis). Indications for scanning were: symptoms/signs of pancreatitis, amylase >3 × ULN, lipase >5 × ULN, positive peritoneal signs, and/or fever of unknown origin.

^vAssessment made difficult by altered levels of consciousness in some/all patients.

^wUnclear if this patient subset was also included in the other studies by this author (Gullo).²⁰,²¹

^xOf the 29 DKA patients with normal pancreatic enzymes, data on abdominal pain and/or vomiting were available for only 21.

^yNo patient with persistently elevated amylase or lipase had CT features of pancreatitis; one patient had US features of pancreatitis; in total, five of 17 were clinically diagnosed as having pancreatitis.

^zFive of 17 patients were clinically diagnosed as having pancreatitis (despite normal CT scans) according to failed enteral feeding + concomitant raised pancreatic enzymes + abdominal distension + epigastric tenderness; one of these patients had a laparotomy.

^aaSeven of 237 patients at some point had pancreatitis (preceding onset of IBD in two, concomitant with onset of IBD in two, and after onset of IBD in three); two of these had raised pancreatic enzyme(s).

^bbCT scan was performed in patients with lipase >3 × ULN; although 38 patients had lipase higher than this threshold, 18 could not undergo imaging as a result of cardiopulmonary instability (n = 7), discharge from ICU prior to CT (n = 7), death (n = 3) and subsequent withdrawal of consent (n = 1). Seven of 20 had morphological alterations of the pancreas on CT (in one patient only a small fluid collection was detected at the pancreatic tail).

AAA, abdominal aortic aneurysm; CAPD, continuous ambulatory peritoneal dialysis; CD, Crohn's disease; CR, case report; CRF, chronic renal failure; CT, computed tomography; DKA, diabetic ketoacidosis; DM, diabetes mellitus; HD, haemodialysis; ICH, intracranial haemorrhage; MRCP, magnetic resonance cholangiopancreatography; MRCP-S, secretin-enhanced MRCP; MRI, magnetic resonance imaging; NR, not recorded; SAH, subarachnoid haemorrhage; SOD, sphincter of Oddi dysfunction; UC, ulcerative colitis; ULN, upper limit of normal; US, ultrasonography; w/o, without.