New-Onset Diabetes Mellitus After the First Attack of Acute Pancreatitis: A Systematic Review and Meta-Analysis

AUTHORS

Cunliang Hu 1 , Qiuping Liu 2 , Niwei Chen 1 , *

1 Department of Gastroenterology, Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, Shanghai, China

2 Department of Respiratory Medicine, Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, Shanghai, China

How to Cite: Hu C, Liu Q, Chen N. New-Onset Diabetes Mellitus After the First Attack of Acute Pancreatitis: A Systematic Review and Meta-Analysis, Iran Red Crescent Med J. 2019 ; 21(7):e91740. doi: 10.5812/ircmj.91740.

ARTICLE INFORMATION

Iranian Red Crescent Medical Journal: 21 (7); e91740
Published Online: July 29, 2019
Article Type: Systematic Review
Received: March 19, 2019
Revised: June 14, 2019
Accepted: June 16, 2019
Crossmark

Crossmark

CHEKING

READ FULL TEXT
Abstract

Context: New-onset diabetes mellitus (DM) after the first attack of acute pancreatitis (AP) has not been fully elucidated.

Objectives: The study aimed to explore the incidence and time-course of pancreatic endocrine insufficiency in patients with new-onset prediabetes or DM after the first attack of AP.

Data Sources: A comprehensive literature review was conducted by searching four major biomedical journal databases (PubMed, Embase, Cochrane Library, and Web of Science).

Study Selection: We included all prospective clinical studies that investigated the change in the metabolization of glucose after hospital discharge following the first attack of AP.

Data Extraction: After quality assessment, data were extracted according to a standard protocol. Because of between-study heterogeneity, data were analyzed by the random-effects method.

Results: The inclusion criteria were met by 12 clinical studies, including 766 patients with the first attack of AP. Prediabetes and/or DM was observed in 51% (95% CI: 55% to 63%) of the patients after the first attack of AP. The pooled incidence of prediabetes and DM after AP was 23% (95% CI: 16% to 30%) and 18% (95% CI: 11% to 26%), respectively. The risk of new-onset prediabetes and DM significantly increased in 1 - 3 years (relative risk (RR): 4.00 (95% CI: 1.68 - 9.53)) and 3 - 5 years (RR: 2.12 (95% CI: 1.9 - 3.8)), respectively.

Conclusions: New-onset prediabetes and/or DM after the first attack of AP developed in 51% of the patients after hospital discharge and the risk of DM increased more than two folds over three years.

Keywords

Acute Pancreatitis Diabetes Mellitus Exocrine Pancreatic Insufficiency Glucose Incidence Islets of Langerhans Meta-Analysis New-Onset Pancreatic Diseases Prediabetic State

Copyright © 2019, Author(s). This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/) which permits copy and redistribute the material just in noncommercial usages, provided the original work is properly cited

1. Context

Diabetes mellitus (DM) is a metabolic disease with a globally increasing prevalence. In 2015, 30.2 million people (9.4%) of the US population had diabetes (1). Diabetes is the seventh leading cause of death in the United States (1). DM is a risk factor for various diseases including coronary heart disease (2), stroke (3), kidney failure (4), lower limb ischemia (5), blindness (6), amputations of the legs and feet (7), and even early death. As for the health and economic perspective, DM is a life-threatening disease, affecting millions of people around the world every year and imposing a large economic burden, with high treatment cost. Prediabetes is a warning sign that people are on the path of diabetes. Impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) are the two forms of prediabetes. The burden of prediabetes is similar to that of DM. The USA estimated 84.1 million US adults with prediabetes in 2015 (1). It is the predisposing stage before DM in which people are more likely to develop diabetes (1, 8-10). About 20% of patients with prediabetes progress to diabetes over five years (8). Moreover, the global burden of prediabetes and DM is expected to increase as the population ages, and nonfatal outcomes require more resources from health care systems. Hence, prevention, and timely diagnosis of prediabetes and DM have important implications. Screening of individuals at high risk of developing DM is needed to identify patients with prediabetes and take measurements to prevent the beginning of DM (11).

Acute pancreatitis (AP) is the most common pancreatic disease (12), with a prevalence reported from 17 countries across Europe, ranging from 4.6 to 100 per 100,000 population (13). It is an acute inflammatory disease of the pancreas characterized by a sudden onset of upper abdominal pain, nausea, emesis, and increased pancreatic digestive enzymes in the serum and urine. Impaired function of the endocrine pancreas causes prediabetes or DM (14). It is generally believed that hyperglycemia can completely recover in most patients after AP. That is why glucose metabolism is not continually measured after leaving the hospital. However, some studies demonstrated that hyperglycemia persists in a proportion of patients after AP (15, 16). Therefore, it is necessary to clarify the time course of prediabetes and DM after the first attack of AP.

There is one published meta-analysis (17) investigating the association between new-onset DM and the first attack of AP. In recent five years, new studies have demonstrated the occurrence of new-onset DM after the first attack of AP. Furthermore, the incidence of DM after AP has changed over recent years. Therefore, it is important to estimate the new-onset DM after the first attack of AP again.

The aim of the study was to perform a systematic review and meta-analysis of studies reporting the incidence and time course of new-onset prediabetes and/or DM after the first attack of AP.

2. Objectives

The study aimed to explore the incidence and time course of pancreatic endocrine insufficiency in patients with new-onset prediabetes or DM after the first attack of AP.

3. Data Sources

We performed a search to evaluate all prospective clinical trials that investigated the change in the metabolization of glucose after hospital discharge following the first attack of AP. We searched four major databases (PubMed, Embase, Cochrane Library, and Web of Science) from the time of inception (1946 for PubMed, 1980 for Embase, 1980 for Cochrane Library, and 1980 for Web of Science) until October 2018. The search strategy used was as follows:

“acute pancreatitis” AND (“diabetes mellitus” OR “type 2 diabetes mellitus” OR “type 1 diabetes mellitus” OR “non-insulin dependent diabetes mellitus” OR “insulin- dependent diabetes mellitus” OR “prediabetic state” OR “impaired glucose tolerance” OR “impaired fasting glucose” OR “impaired glucose regulation”).

The references of all the included studies were also searched for other relevant articles. Open Grey and ClinicalTrials.gov were also searched. The search was limited to publications in English.

4. Study Selection

The criteria for including studies are in the following: (1) prospective clinical human studies; (2) patients aged > 18 years experiencing the first attack of AP; (3) reporting diagnostic laboratory testing for pancreatic endocrine insufficiency, and (4) at least one-month follow-up time after hospital discharge.

Studies were excluded if any of the following existed: (1) patients had pre-existing DM or prediabetes, or studies did not state whether patients had pre-existing DM or prediabetes; (2) patients performed pancreatic surgery (such as necrosectomy or pancreatic resection), and (3) patients with autoimmune, chronic or hereditary pancreatitis and gestational diabetes.

5. Data Extraction

To ensure accuracy, two authors (CunLiang Hu and QiuPing Liu) independently performed data extraction, study selection, and quality assessment and a senior author (NiWei Chen) reviewed the process. Disagreements were resolved by discussion between the two authors and the senior author. The following information was obtained from each selected study: (1) study/author, (2) country, (3) study design, (4) follow-up time, (5) severity of AP, (6) endocrine function tests, (7) the number of patients under endocrine assessment, (8) the number of multiple attacks of AP at follow-up, (9) prediabetes/DM after AP, (10) diabetic diet and physical exercise, DM treated with oral diabetes drugs, and DM treated with insulin after AP, (11) total individuals studied, (12) age, (13) the proportion of male to female patients, (14) body mass index (BMI), (15) etiology of AP (biliary, hyperlipemia, alcohol, other), (16) the number of participating centers, and (17) criteria used to diagnose AP and classify its severity. The corresponding authors were contacted if necessary.

5.1. Quality Assessment

We used the Newcastle-Ottawa Scale (NOS) to evaluate the quality of eligible studies (18). According to the three aspects of research design (patient selection, comparability of research groups, and exposure/outcomes of research participants), it allocates up to nine points to each study. If the score is 5 or more, the quality of research is considered high; if the score is less than 5, the quality of research is considered low (19).

5.2. Definitions

AP is defined as having two of the following three items: (1) abdominal pain characteristic of AP, (2) serum amylase and/or lipase three times the upper limit of normal, and (3) AP characteristic findings on computed tomography (20). Fasting blood glucose (FBG) and/or 2-h oral glucose tolerance test (OGTT) were used to define prediabetes and DM. IFG is defined as FBG ≥ 5.6 mmol/L and < 7.0 mmol/L, or ≥ 6.1 mmol/L and < 7.0 mmol/L. IGT is defined as 2-h OGTT ≥ 7.8 mmol/L and < 11.1 mmol/L. Diabetes is defined by typical diabetes symptoms with any of the following items: (1) FBG ≥ 7.0 mmol/L, (2) random blood glucose ≥ 11.1 mmol/L, and (3) 2-h OGTT ≥ 11.1 mmol/L (21, 22).

5.3. Data Processing and Statistical Analysis

The pooled prevalence for each of the outcomes was calculated. Random-effects model, pooled prevalence, and associated 95% confidence intervals (CIs) were used to evaluate pancreatic endocrine dysfunction. R console version 3.5.1 was used to perform forest plot analysis, funnel plot analysis, sensitivity analysis, subgroup analysis, and meta-regression analysis. Between-study heterogeneity in different subgroups was tested using the I2-statistics. The I2 values of < 25%, 25% - 50%, 50% - 75%, and ≥ 75% were classified as low, moderate, high, and very high heterogeneity, respectively (23). Egger’s test was used to assess publication bias. Sensitivity analyses were limited to studies that used the 1999 WHO definition of DM and studies with SAP. Subgroup analysis was performed to find a possible change in pancreatic endocrine dysfunction over time by comparing studies of different follow-up periods (up to 12 months, 12 - 36 months, 36 - 60 months, and more than 60 months). Meta-regression analysis was conducted to explore whether sex, age, follow-up time, and etiology of AP (proportion of biliary, hyperlipemia, alcohol, other etiology) had an important impact on prognosis. The review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (24).

6. Results

6.1. Study Characteristics

The search retrieved a total number of 4834 studies of which, 65 relevant papers were included in the assessment. Finally, 12 studies were used in the analysis (Figure 1) (16, 25-35). Cohen’s kappa coefficient was 0.773 (95% CI: 0.610 - 0.936), which indicates a strong consistency. Table 1 lists the characteristics of the included studies. The included studies recruited 766 AP patients in total (Table 2). All the studies performed prospective cohorts. Eight studies (16, 27, 29, 30, 32-35) were conducted in Europe, three (25, 26, 31) in Asia, and one (28) in South America. All studies were conducted in single centers.

Table 1. Characteristics of the Included Studies
StudyYearCountryStudy DesignFollow-up Time (Mean Unless Specified), moSeverity of Acute Pancreatitis, %Test of Endocrine Function UsedTotal No. of Individuals Undergoing Endocrine AssessmentPresence of Multiple Attacks of AP at Follow-up, %Prediabetes, %DM, %Diabetic Diet and Physical Exercise, %DM Treated with Oral Diabetes Drugs, %DM Treated with Insulin, %
Tu (25)2018ChinaProspective cohort42.93MAP 54 (21), MSAP 42 (16.4), SAP 160 (62.5)OGTT (FBG, 2hPG), FINS, HBA1c, HOMA-β,HOMA-IR256Not statedIGT 72 (28.1) (from author)82 (32.0) (from author)Not statedNot statedNot stated
Tu (26)2017ChinaProspective cohort42.93MAP 10 (8.8), MSAP 12 (10.6), SAP 91 (80.6)OGTT (FBG, 2hPG), FINS, HBA1c, HOMA-β,HOMA-IR113Not statedIGT 33 (29.2)34 (30.1)Not statedNot statedNot stated
Nikkola (27)2017FinlandProspective cohort126MAP 35 (74.5), MSA and SAP, 12 (25.5)HbA1c, OGTT, FPG, glucagon C-peptide test4714 (29.8)13 (27.7)13 (27.7)Not statedNot statedNot stated
Winter Gasparoto (28)2015BrazilProspective cohort34.8All SAPOGTT, C-peptide, HOMA14Not stated7 (58.3)5 (41.7)Not statedNot statedNot stated
Vujasinovic (29)2014SloveniaProspective cohort32.4MAP 67 (67.0), MSAP 15 (15.0), SAP 18 (18.0)HbA1c, OGTT75Not statedNot stated6 (8.0) (from author)Not statedNot statedNot stated
Nikkola (30)2013FinlandProspective cohort61.8Not statedFBG, HbA1c, OGTT, C-peptide150 (0)1 (6.7)0 (0)Not statedNot statedNot stated
Wu (31)2011ChinaProspective cohort42MAP 24 (40.7), SAP 35 (59.7)FBG, HbA1c, FBI, C-peptide, HOMA59Not statedNot statedNot statedNot statedNot statedNot stated
Uomo (32)2010ItalyProspective cohort179.5 (median)All SAPFBG, OGTT38Not statedNot stated6 (15.9)Not statedNot statedNot stated
Pelli (33)2009FinlandProspective cohort47 (median )MAP 41 (75), SAP 13 (25)FBG, HbA1c, OGTT46one time 10 (19), two times 1 (18.5), three times 1 (18.5), (baseline 54)IFG 7 (15.2), IGT 5 (10.9)5 (10.9)Not statedNot statedNot stated
Boreham (16)2003UKProspective cohort3MAP 16 (69.6), SAP 7 (30.4)FPG23No, only a single attackNot stated4 (17.4)Not stated1 (4.3)Not stated
Ibars (34)2002SpainProspective cohort1, 6 and 12MAP 45 (71.0), SAP 18 (28.0)OGTT, arginine test55Not stated7 (12.7)6 (10.9)Not statedNot statedNot stated
Doepel (35)1993FinlandProspective cohort74.4All SAPBG, C-peptide, HbA1, HbA1c, OGGT24Not statedIGT 4 (16.7)7 (29.2)4 (16.7)1 (4.2)2 (8.3)
Table 2. Baseline Characteristics of the Included Studiesa
StudyYearTotal Individuals Studied, No.Age, yGenderBMI, kg/m2Etiology
MaleFemaleBiliaryHyperlipemiaAlcoholOther
Tu (25)201825655168 (65.6)88 (34.4)Not stated147 (57.5)88 (34.5)7 (2.7)14 (5.3)
Tu (26)20171134675 (66.4)38 (33.6)< 18.5 (4.4% of patients)65 (57.5)39 (34.5)3 (2.7)6 (5.3)
Nikkola (27)2017774869 (90)8 (10)27.7Not statedNot statedNot statedNot stated
Winter Gasparoto (28)201516489 (56.2)7 (43.8)Not stated10 (62.5)2 (12.5)4 (25.0)0
Vujasinovic (29)201410056.565 (65.0)35 (35.0)No stated36 (36.0)6 (6.0)42 (42.0)16 (16.0)
Nikkola (30)2013184718 (100)029.60018 (100)0
Wu (31)20115958.833 (56)26 (44)23.242 (71)7 (12)7 (12)3 (5)
Uomo (32)20104063.517 (42.5)23 (57.5)one:29; another: 3228 (70.0)5 (12.5)07 (17.5)
Pelli (33)2009544947 (87.0)7 (13.0)2754 (100)000
Boreham (16)2003235513 (56.6)10 (43.4)Not stated13 (56.6)1 (4.3)5 (21.7)4 (17.4)
Ibars (34)20026362.317 (27)46 (73)Not stated63 (100)000
Doepel (35)1993375225 (67.6)12 (32.4)Not stated3 (8.1)028 (75.7)6 (16.2)

Abbreviation: BMI, body mass index.

aValues are expressed as No. (%) unless otherwise indicated.

Flowchart of the study selection process
Figure 1. Flowchart of the study selection process

6.2. Quality of Studies

Using the NOS, we assessed the quality of the included studies (Table 3) (18). All of the 12 studies had high quality.

Table 3. The Methodological Quality of Studies Stratified by Study Designa
StudyYearMeasured Items on the Newcastle-Ottawa ScaleScores (High/Low Quality)
Patient SelectionComparabilityExposure/Outcome
Cohort(1) Representativeness of the exposed cohort, (2)Selection of non-exposed cohort, (3)Ascertainment of exposure, (4)Demonstration that the outcome of interest was not present at the start of the study(1) Comparability of cohorts on the basis of the design or analysis (maximum two points)(1) Assessment of outcome, (2) Follow-up was long enough for outcomes to occur, (3) Adequacy of follow-up of cohortslow quality: 0-4, high quality: 5 - 9
Tu (25)2018(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP, (4)a Outcome of interest was not present at the start of study(1)a Study controls for severity of AP(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for8
Tu (26)2017(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP, (4)a Outcome of interest was not present at the start of study(1)a Study controls for severity of AP(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for8
Nikola (27)2017(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP(1)a Study controls for severity of AP and etiology(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for7
Winter Gasparoto (28)2015(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort (3)a Prospective recruitment of patients who met the criteria for AP(1)a Study controls for severity of AP(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for7
Vujasinovic (29)2014(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP, (4)a Outcome of interest was not present at the start of study(1)a Study controls for severity of AP(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for8
Nikola (30)2013(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP, (4)a Outcome of interest was not present at the start of study(1)a Study controls for severity of AP and etiology(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for8
Wu (31)2011(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP, (4)a Outcome of interest was not present at the start of study(1)a Study controls for severity of AP(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for,8
Uomo (32)2010(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP(1)a Study controls for severity of AP(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a Subjects lost to follow up unlikely to introduce bias - description provided on those who lost7
Pelli (33)2009(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP(1)a Study controls for severity of AP and etiology(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for7
Boreham (16)2003(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP, (4)a Outcome of interest was not present at the start of study(1)a Study controls for severity of AP(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a All subjects were accounted for8
Ibars (34)2002(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Ascertainment of exposure (diagnosis of AP) was through secure surgical records(1)a Study controls for severity of AP and etiology(1)a Outcome was assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a Subjects lost to follow up unlikely to introduce bias - description provided on those who lost7
Doepel (35)1993(1)a Truly representative of patients with AP, (2)a Endocrine assessment was done for patients selected from exposed cohort, (3)a Prospective recruitment of patients who met the criteria for AP(1)a Study controls for severity of AP(1)a Outcome assessed by independent endocrine testing, (2)a Follow-up was long enough for outcome to occur, (3)a Subjects lost to follow up unlikely to introduce bias - description provided on those who lost7

Abbreviation: AP, acute pancreatitis.

aPoints on the scale.

6.3. Publication Bias

Publication bias was examined and there was no evidence of funnel plot asymmetry for DM studies with a P value of 0.09 (Egger’s test). Because the number of included studies was less than 10 for prediabetes and/or DM and prediabetes studies, we did not assess the publication bias for them.

6.4. Prediabetes and/or Diabetes

There were eight studies reporting on prediabetes and/or DM after AP (25-28, 30, 33-35), with 517 patients included. The pooled prevalence of prediabetes and/or DM was 51% (95% CI: 41% to 63%), with high statistical heterogeneity between studies (I2 = 82%). Table 4 and Figure 2A show the time course of prediabetes and/or DM in the included studies.

Table 4. Subgroup Analysis According to the Follow-up Time
Follow-up Time, moPrediabetes and/or DMPrediabetesDM
No. of StudiesTotal No. of PatientsNo. of Patients with Prediabetes and/or DMPooled Prevalence, (95% CI)I2, (%)Relative Risk, (95% CI)aNo. of StudiesNo. of StudiesNo. of Patients with PrediabetesPooled Prevalence, (95% CI)I2, (%)Relative Risk, (95% CI)aNo. of StudiesTotal No. of PatientsNo. of Patients with DMPooled Prevalence, (95% CI)I2, (%)Relative risk (95% CI)a
≤ 121 (34)56130.23 [0.14; 0.37]bNot applicable1 (34)5670.13 [0.06; 0.25]bNot applicable2 (16, 34)79100.13 [0.07; 0.23]01.03 [0.51; 2.09]
12 - 361 (28)14120.86 [0.69; 1.00]b3.69 [1.69; 6.22]1 (28)1470.50 [0.30; 0.84]b4.00 [1.68; 9.53]2 (28, 29)89110.17 [0.04; 0.74]871.35 [0.30; 6.11]
36 - 603 (25, 26, 33)4152380.55 [0.47; 0.66]602.28 [1.69; 3.09]3 (25, 26, 33)4151170.28 [0.24; 0.33]02.24 [1.44; 3.50]3 (25, 26, 33)4151210.27 [0.19; 0.38]682.12 [1.42; 3.16]
> 603 (27, 30, 35)86380.45 [0.28; 0.73]671.74 [0.86; 3.55]3 (27, 30, 35)86180.21 [0.12; 0.37]271.57 [0.84; 2.93]4 (27, 30, 32, 35)124260.23 [0.15; 0.35]261.63 [1.02; 2.59]

Abbreviation: DM, diabetes mellitus.

aCompared to follow-up ≤ 12 months after the first attack of AP.

bNot enough number of studies.

The pooled prevalence of pancreatic endocrine insufficiency with time after the first attack of AP. A, prediabetes and/or diabetes mellitus; B, prediabetes; C, diabetes mellitus. 1, ≤ 12 months; 2, 12 - 36 months; 3, 36 - 60 months; 4, &gt; 60 months.
Figure 2. The pooled prevalence of pancreatic endocrine insufficiency with time after the first attack of AP. A, prediabetes and/or diabetes mellitus; B, prediabetes; C, diabetes mellitus. 1, ≤ 12 months; 2, 12 - 36 months; 3, 36 - 60 months; 4, > 60 months.

6.5. Prediabetes

All the eight studies (25-28, 30, 33-35) reported on new-onset prediabetes after AP, comprising 571 patients with a pooled prevalence of 23% (95% CI: 16% to 30%). There was moderate statistical heterogeneity between studies (I2 = 69%) (Figure 3A). Sensitivity analysis in individuals with severe AP (SAP) included two studies (28, 35), with 38 patients involved. The pooled prevalence of prediabetes was 31% (95% CI: 11% to 90%) and heterogeneity between the studies did not decrease (I2 = 77%). Sensitivity analysis involved studies that used the 1999 WHO definitions (21), including three studies (25, 26, 33) with 415 patients. The pooled prevalence was 28% (95% CI: 24% to 33%), with no statistical heterogeneity between studies (I2 = 0%). Subgroup analysis involved follow-up time as shown in Table 2. The prevalence of prediabetes after AP stayed within the range of 13% - 50% and the relative risk of developing prediabetes at any time-point compared to up to 12 months initially decreases with time (Table 4). Time-course analysis of the prevalence of prediabetes is presented in Figure 2B.

The pooled prevalence of pancreatic endocrine insufficiency after the first attack of AP
Figure 3. The pooled prevalence of pancreatic endocrine insufficiency after the first attack of AP

6.6. Diabetes Mellitus

There were 11 studies (16, 25-30, 32-35) reported on new-onset DM. They included 707 patients with a pooled diabetes prevalence of 18% (95% CI: 11% to 26%). There was high statistical heterogeneity between studies (I2 = 82%) (Figure 3B). Sensitivity analysis included individuals with SAP in three studies (28, 32, 35) involving 707 patients. The pooled prevalence of DM in patients with SAP was 26% (95% CI: 16% to 41%), with decreased heterogeneity between the studies (I2 = 26%). Sensitivity analysis involved studies that used the 1999 WHO definitions, including three studies (25, 26, 33) with 415 patients. The pooled prevalence was 28% (95% CI: 24% to 33%), with no heterogeneity between the studies (I2 = 0%). Subgroup analysis, according to the follow-up time, is shown in Table 2. The prevalence of DM after AP remained within the range of 13% - 27%. There was a 2.12-fold increased risk of new-onset DM at 36 - 60 months, which was statistically significant (Table 4). Time-course analysis of the prevalence of DM is presented in Figure 2C.

6.7. Treatment of DM

Limited studies described the treatment of DM (including diabetic diet, physical exercise, treatment with oral diabetes drugs, and treatment with insulin); thus, we did not assess the pooled prevalence of them.

6.8. Meta-Regression Analysis

Meta-regression was performed using the following moderators: age, follow-up time, the proportion of males, the proportion of biliary etiology, the proportion of hyperlipemia etiology, the proportion of alcohol etiology, the proportion of other etiologies, and the total number of patients. The analysis showed that the prevalence of prediabetes and/or DM, prediabetes, and DM was not significantly influenced by these factors (Table 5).

Table 5. Results of Meta-Regression Analysisa
AgeP Values for the Covariates Tested
Follow-up TimeMale, %Biliary, %Hyperlipemia, %Alcohol, %Other Etiologies, %Total No. of Patients
Prediabetes and/or diabetes mellitus0.080.470.200.550.280.370.360.38
Prediabetes0.170.770.370.490.280.420.910.41
Diabetes mellitus0.150.620.120.980.160.470.550.73

aSome covariates unavailable were excluded from the model.

7. Discussion

We systematically evaluated all 12 available clinical studies reporting on patients with prediabetes and/or DM after the first attack of AP who had been discharged from the hospital. The pooled prevalence of prediabetes and/or DM after AP was 51% while prediabetes and DM were observed in 23% and 18% of the individuals after AP, respectively. Furthermore, the occurrence of prediabetes or DM after AP was more frequent when pancreatitis was more severe; such patients had a higher prevalence of both prediabetes (31%) and DM (26%). The meta-regression analysis indicated that the risk of prediabetes or DM after AP was irrespective of patients’ age, gender, etiology, the total number of patients, and follow-up duration. The results also indicated that prediabetes and DM appeared very early after AP. The prevalence of prediabetes and DM after AP was 13% for both in the first 12 months. Unfortunately, no study in this review followed patients with prediabetes to confirm whether they developed DM, but studies proposed that patients with prediabetes are more likely to develop DM (1, 8-10). Compared to the previous meta-analysis (17), we found that the prevalence of new-onset DM after the first attack of AP increases from 37% to 51%. Hence, closer follow-ups of patients after AP, screening of patients with prediabetes, and taking measures to prevent the beginning of DM are necessary.

Pancreatogenic DM is an acknowledged condition classified as type 3c diabetes (36-38). Pancreatic necrosis can lead to a decline in the number of β-cell and insulin secretion, known to be the main reason for DM development after AP. The pooled prevalence of prediabetes or DM in patients with SAP was 31% and 26%, respectively. Hence, patients with SAP are more likely to develop prediabetes or DM. First, patients with SAP suffer from a larger decrease in β-cell shortly, with more loss of functional reserve capacity of the gland (15, 16, 26). Second, compared to mild acute pancreatitis (MAP) and moderately severe acute pancreatitis (MSAP), SAP patients with a higher percentage of necrosis are more likely to perform necrosectomy (39), with more number of β-cell decrement. Third, some metabolic factors, such as obesity and hypertriglyceridemia as the risk factors for SAP, contribute to prediabetes or DM after AP (40-44). It is also worth mentioning that pre-existing DM itself places individuals at greater risk of developing DM after AP (1, 9). This review excluded patients undergoing necrosectomy and pre-existing DM.

We showed an increasing trend in the prevalence of prediabetes and diabetes as follow-up time increases. Age and recurrent attacks of AP during the follow-up time may have effects on this trend. However, meta-regression analysis showed that age was not a significant influencing factor. Unfortunately, just four studies reported recurrent attacks of AP and only one study made a specific notion of recurrent attacks. Due to insufficient data, this was not analyzed here in this study. Remarkable is that the highest prevalence of prediabetes and diabetes occurred in 1 - 3 years and 3 - 5 years after AP, respectively; however, it decreased thereafter. The question is that did part of prediabetes decreases turn into diabetes? Unfortunately, no study in this review followed patients with prediabetes to confirm whether they developed DM. Most importantly, the prevalence of prediabetes and diabetes decreased after 3 and 5 years of follow-up, respectively. The change in increasing and decreasing trends in the prevalence over time between prediabetes and DM is worthy of attention. It suggests that the human endocrine pancreas has a certain regenerative capacity after acute pancreatitis. This is supported by human and animal experimental models (45). The possibility that pancreatic endocrine function changes with time after acute pancreatitis deserves doctors’ attention in the long-term follow-up of acute pancreatitis.

The advantage of this systematic review is that we followed a comprehensive strategy allowing to include all 766 individuals with the first attack of AP after hospital discharge from prospective studies. Open Grey and ClinicalTrials.gov were also searched. Patients with pre-existing DM and pancreatic surgery were excluded, allowing to report new-onset prediabetes and DM after AP accurately. Furthermore, we contacted authors for missing data or unreported variables. By doing so, we included two more studies (25, 29). Statistical analysis is robust when the random-effects model is used to provide the most conservative estimates. We also conducted sensitivity and subgroup analyses (in order to explore possible grounds for statistical heterogeneity) and meta-regression (in order to evaluate the effect of potential confounders).

Attention should be paid to the limitations of this study. First, the high heterogeneity between studies, found in sensitivity and subgroup analyses, limited between-study comparability and indicated that meta-analysis results should be interpreted with caution. Differences in pancreatic endocrine function assessment methods, study designs, and geographical locations may have been the causes of high heterogeneity. Second, there were limited studies in sensitivity analysis and subgroup analysis. Hence, the results of these analyses also should be interpreted with caution. Third, not all studies mentioned BMI, recurrent AP, and DM treatment. Due to insufficient data, it was impossible to explain the impact of these factors. Fourth, the results of the time adjustment analysis should be carefully explained. Our data came from multiple studies with different follow-up times, rather than longitudinal studies with repeated measurements at different time-points. A prospective study of a large cohort of patients over several years is more appropriate for this purpose. Fifth, although this study excluded patients who underwent surgery, they may be included because some studies did not mention surgery information. It has been reported that distal pancreatectomy puts AP patients at a higher risk of DM (46). Sixth, by excluding non-English studies, language bias may have existed. Finally, the management of AP after discharge from the hospital may affect pancreatic endocrine function but, due to insufficient data, this possible confounding factor could not be adjusted for.

We hope to provide some suggestions for future studies on endocrine dysfunction in AP follow-up. First, studies need to declare whether patients had pre-existing DM or underwent necrosectomy. Second, the severity of the disease should be well described, preferably classified according to recognized diseases (e.g., the revised Atlanta Classification (47)). Moreover, BMI, recurrent AP, and DM treatment should also be mentioned in future studies. The consequences of pancreatic endocrine insufficiency may be a major burden on AP patients’ lives. Doctors should know which patients are at risk.

In conclusion, this systematic literature review confirms that prediabetes and DM are common after the first attack of AP. It is also suggested that pancreatic endocrine function recovers with time after AP; hence, a formal follow-up of patients after AP is important. Further research should pay more attention to the pathogenesis, detection, and screening of prediabetes/DM after the first attack of AP.

Footnotes

References

  • 1.

    Centers for Disease Control and Prevention . Diabetes Report Card 2017. 2017. Available from: https://www.cdc.gov/diabetes/pdfs/library/diabetesreportcard2017-508.pdf.

  • 2.

    Aronson D, Edelman ER. Coronary artery disease and diabetes mellitus. Heart Fail Clin. 2016;12(1):117-33. doi: 10.1016/j.hfc.2015.08.010. [PubMed: 26567979].

  • 3.

    Bell DS. Stroke in the diabetic patient. Diabetes Care. 1994;17(3):213-9. doi: 10.2337/diacare.17.3.213. [PubMed: 8174450].

  • 4.

    Gross JL, de Azevedo MJ, Silveiro SP, Canani LH, Caramori ML, Zelmanovitz T. Diabetic nephropathy: Diagnosis, prevention, and treatment. Diabetes Care. 2005;28(1):164-76. doi: 10.2337/diacare.28.1.164. [PubMed: 15616252].

  • 5.

    Siitonen OI, Niskanen LK, Laakso M, Siitonen JT, Pyorala K. Lower-extremity amputations in diabetic and nondiabetic patients. A population-based study in eastern Finland. Diabetes Care. 1993;16(1):16-20. doi: 10.2337/diacare.16.1.16. [PubMed: 8422771].

  • 6.

    Cheung N, Mitchell P, Wong TY. Diabetic retinopathy. Lancet. 2010;376(9735):124-36. doi: 10.1016/S0140-6736(09)62124-3. [PubMed: 20580421].

  • 7.

    Jeffcoate WJ, Harding KG. Diabetic foot ulcers. Lancet. 2003;361(9368):1545-51. doi: 10.1016/S0140-6736(03)13169-8. [PubMed: 12737879].

  • 8.

    DeJesus RS, Breitkopf CR, Rutten LJ, Jacobson DJ, Wilson PM, Sauver JS. Incidence rate of prediabetes progression to diabetes: Modeling an optimum target group for intervention. Popul Health Manag. 2017;20(3):216-23. doi: 10.1089/pop.2016.0067. [PubMed: 27689627].

  • 9.

    International Diabetes Federation. IDF diabetes atlas. 2017. Available from: http://diabetesatlas.org/IDF_Diabetes_Atlas_8e_interactive_EN/.

  • 10.

    Tabak AG, Herder C, Rathmann W, Brunner EJ, Kivimäki M. Prediabetes: a high-risk state for diabetes development. Lancet. 2012;379(9833):2279-90. doi: 10.1016/S0140-6736(12)60283-9. [PubMed: 22683128]. [PubMed Central: PMC3891203].

  • 11.

    Diabetes Prevention Program Research G, Knowler WC, Fowler SE, Hamman RF, Christophi CA, Hoffman HJ, et al. 10-year follow-up of diabetes incidence and weight loss in the Diabetes Prevention Program Outcomes Study. Lancet. 2009;374(9702):1677-86. doi: 10.1016/S0140-6736(09)61457-4. [PubMed: 19878986]. [PubMed Central: PMC3135022].

  • 12.

    Lankisch PG, Apte M, Banks PA. Acute pancreatitis. Lancet. 2015;386(9988):85-96. doi: 10.1016/S0140-6736(14)60649-8. [PubMed: 25616312].

  • 13.

    Roberts SE, Morrison-Rees S, John A, Williams JG, Brown TH, Samuel DG. The incidence and aetiology of acute pancreatitis across Europe. Pancreatology. 2017;17(2):155-65. doi: 10.1016/j.pan.2017.01.005. [PubMed: 28159463].

  • 14.

    Pendharkar SA, Asrani VM, Xiao AY, Yoon HD, Murphy R, Windsor JA, et al. Relationship between pancreatic hormones and glucose metabolism: A cross-sectional study in patients after acute pancreatitis. Am J Physiol Gastrointest Liver Physiol. 2016;311(1):G50-8. doi: 10.1152/ajpgi.00074.2016. [PubMed: 27173509].

  • 15.

    Andersson B, Pendse ML, Andersson R. Pancreatic function, quality of life and costs at long-term follow-up after acute pancreatitis. World J Gastroenterol. 2010;16(39):4944-51. doi: 10.3748/wjg.v16.i39.4944. [PubMed: 20954281]. [PubMed Central: PMC2957603].

  • 16.

    Boreham B, Ammori BJ. A prospective evaluation of pancreatic exocrine function in patients with acute pancreatitis: Correlation with extent of necrosis and pancreatic endocrine insufficiency. Pancreatology. 2003;3(4):303-8. doi: 10.1159/000071768. [PubMed: 12890992].

  • 17.

    Das SL, Singh PP, Phillips AR, Murphy R, Windsor JA, Petrov MS. Newly diagnosed diabetes mellitus after acute pancreatitis: A systematic review and meta-analysis. Gut. 2014;63(5):818-31. doi: 10.1136/gutjnl-2013-305062. [PubMed: 23929695].

  • 18.

    Wells GA, Shea B, O'Connell D. The newcastle-ottawa scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. 2019. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp.

  • 19.

    Halpin SJ, Ford AC. Prevalence of symptoms meeting criteria for irritable bowel syndrome in inflammatory bowel disease: Systematic review and meta-analysis. Am J Gastroenterol. 2012;107(10):1474-82. doi: 10.1038/ajg.2012.260. [PubMed: 22929759].

  • 20.

    Dellinger EP, Forsmark CE, Layer P, Levy P, Maravi-Poma E, Petrov MS, et al. Determinant-based classification of acute pancreatitis severity: An international multidisciplinary consultation. Ann Surg. 2012;256(6):875-80. doi: 10.1097/SLA.0b013e318256f778. [PubMed: 22735715].

  • 21.

    World Health Organization. Definition, diagnosis and classification of diabetes mellitus and its complicationsm report of a WHO consultation. 1999. Available from: http://apps.who.int/iris/bitstream/handle/10665/66040/WHO_NCD_NCS_99.2.pdf?sequence=1&isAllowed=y.

  • 22.

    American Diabetes A. 2. Classification and diagnosis of diabetes: Standards of medical care in diabetes-2018. Diabetes Care. 2018;41(Suppl 1):S13-27. doi: 10.2337/dc18-S002. [PubMed: 29222373].

  • 23.

    Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557-60. doi: 10.1136/bmj.327.7414.557. [PubMed: 12958120]. [PubMed Central: PMC192859].

  • 24.

    Moher D, Liberati A, Tetzlaff J, Altman DG, Prisma Group . Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. J Clin Epidemiol. 2009;62(10):1006-12. doi: 10.1016/j.jclinepi.2009.06.005. [PubMed: 19631508].

  • 25.

    Tu J, Yang Y, Zhang J, Yang Q, Lu G, Li B, et al. Effect of the disease severity on the risk of developing new-onset diabetes after acute pancreatitis. Medicine (Baltimore). 2018;97(22). e10713. doi: 10.1097/MD.0000000000010713. [PubMed: 29851776]. [PubMed Central: PMC6392884].

  • 26.

    Tu J, Zhang J, Ke L, Yang Y, Yang Q, Lu G, et al. Endocrine and exocrine pancreatic insufficiency after acute pancreatitis: Long-term follow-up study. BMC Gastroenterol. 2017;17(1):114. doi: 10.1186/s12876-017-0663-0. [PubMed: 29078749]. [PubMed Central: PMC5658961].

  • 27.

    Nikkola J, Laukkarinen J, Lahtela J, Seppanen H, Jarvinen S, Nordback I, et al. The long-term prospective follow-up of pancreatic function after the first episode of acute alcoholic pancreatitis: Recurrence predisposes one to pancreatic dysfunction and pancreatogenic diabetes. J Clin Gastroenterol. 2017;51(2):183-90. doi: 10.1097/MCG.0000000000000564. [PubMed: 28060789].

  • 28.

    Winter Gasparoto RC, Racy Mde C, De Campos T. Long-term outcomes after acute necrotizing pancreatitis: what happens to the pancreas and to the patient? JOP. 2015;16(2):159-66. doi: 10.6092/1590-8577/2958. [PubMed: 25791549].

  • 29.

    Vujasinovic M, Tepes B, Makuc J, Rudolf S, Zaletel J, Vidmar T, et al. Pancreatic exocrine insufficiency, diabetes mellitus and serum nutritional markers after acute pancreatitis. World J Gastroenterol. 2014;20(48):18432-8. doi: 10.3748/wjg.v20.i48.18432. [PubMed: 25561813]. [PubMed Central: PMC4277983].

  • 30.

    Nikkola J, Raty S, Laukkarinen J, Seppanen H, Lappalainen-Lehto R, Jarvinen S, et al. Abstinence after first acute alcohol-associated pancreatitis protects against recurrent pancreatitis and minimizes the risk of pancreatic dysfunction. Alcohol Alcohol. 2013;48(4):483-6. doi: 10.1093/alcalc/agt019. [PubMed: 23538610].

  • 31.

    Wu D, Xu Y, Zeng Y, Wang X. Endocrine pancreatic function changes after acute pancreatitis. Pancreas. 2011;40(7):1006-11. doi: 10.1097/MPA.0b013e31821fde3f. [PubMed: 21705952].

  • 32.

    Uomo G, Gallucci F, Madrid E, Miraglia S, Manes G, Rabitti PG. Pancreatic functional impairment following acute necrotizing pancreatitis: Long-term outcome of a non-surgically treated series. Dig Liver Dis. 2010;42(2):149-52. doi: 10.1016/j.dld.2009.08.007. [PubMed: 19836318].

  • 33.

    Pelli H, Lappalainen-Lehto R, Piironen A, Jarvinen S, Sand J, Nordback I. Pancreatic damage after the first episode of acute alcoholic pancreatitis and its association with the later recurrence rate. Pancreatology. 2009;9(3):245-51. doi: 10.1159/000212089. [PubMed: 19407478].

  • 34.

    Ibars EP, Sanchez de Rojas EA, Quereda LA, Ramis RF, Sanjuan VM, Peris RT. Pancreatic function after acute biliary pancreatitis: Does it change? World J Surg. 2002;26(4):479-86. doi: 10.1007/s00268-001-0253-7. [PubMed: 11910484].

  • 35.

    Doepel M, Eriksson J, Halme L, Kumpulainen T, Hockerstedt K. Good long-term results in patients surviving severe acute pancreatitis. Br J Surg. 1993;80(12):1583-6. doi: 10.1002/bjs.1800801229. [PubMed: 8298931].

  • 36.

    Expert Committee on the D, Classification of Diabetes M. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care. 2003;26 Suppl 1:S5-20. doi: 10.2337/diacare.26.2007.s5. [PubMed: 12502614].

  • 37.

    Ewald N, Kaufmann C, Raspe A, Kloer HU, Bretzel RG, Hardt PD. Prevalence of diabetes mellitus secondary to pancreatic diseases (type 3c). Diabetes Metab Res Rev. 2012;28(4):338-42. doi: 10.1002/dmrr.2260. [PubMed: 22121010].

  • 38.

    Ewald N, Bretzel RG. Diabetes mellitus secondary to pancreatic diseases (Type 3c)--are we neglecting an important disease? Eur J Intern Med. 2013;24(3):203-6. doi: 10.1016/j.ejim.2012.12.017. [PubMed: 23375619].

  • 39.

    van Santvoort HC, Bakker OJ, Bollen TL, Besselink MG, Ahmed Ali U, Schrijver AM, et al. A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome. Gastroenterology. 2011;141(4):1254-63. doi: 10.1053/j.gastro.2011.06.073. [PubMed: 21741922].

  • 40.

    Eckel RH, Grundy SM, Zimmet PZ. The metabolic syndrome. Lancet. 2005;365(9468):1415-28. doi: 10.1016/S0140-6736(05)66378-7. [PubMed: 15836891].

  • 41.

    Martinez J, Johnson CD, Sanchez-Paya J, de Madaria E, Robles-Diaz G, Perez-Mateo M. Obesity is a definitive risk factor of severity and mortality in acute pancreatitis: An updated meta-analysis. Pancreatology. 2006;6(3):206-9. doi: 10.1159/000092104. [PubMed: 16549939].

  • 42.

    Tsuang W, Navaneethan U, Ruiz L, Palascak JB, Gelrud A. Hypertriglyceridemic pancreatitis: Presentation and management. Am J Gastroenterol. 2009;104(4):984-91. doi: 10.1038/ajg.2009.27. [PubMed: 19293788].

  • 43.

    Hitman GA. Metabolic syndrome and future risk of diabetes. Diabet Med. 2011;28(8):883. doi: 10.1111/j.1464-5491.2011.03369.x. [PubMed: 21749440].

  • 44.

    Grundy SM. Metabolic syndrome update. Trends Cardiovasc Med. 2016;26(4):364-73. doi: 10.1016/j.tcm.2015.10.004. [PubMed: 26654259].

  • 45.

    Romer AI, Sussel L. Pancreatic islet cell development and regeneration. Curr Opin Endocrinol Diabetes Obes. 2015;22(4):255-64. doi: 10.1097/MED.0000000000000174. [PubMed: 26087337]. [PubMed Central: PMC4660868].

  • 46.

    Scavini M, Dugnani E, Pasquale V, Liberati D, Aleotti F, Di Terlizzi G, et al. Diabetes after pancreatic surgery: Novel issues. Curr Diab Rep. 2015;15(4):16. doi: 10.1007/s11892-015-0589-2. [PubMed: 25702096].

  • 47.

    Banks PA, Bollen TL, Dervenis C, Gooszen HG, Johnson CD, Sarr MG, et al. Classification of acute pancreatitis--2012: Revision of the Atlanta classification and definitions by international consensus. Gut. 2013;62(1):102-11. doi: 10.1136/gutjnl-2012-302779. [PubMed: 23100216].

  • COMMENTS

    LEAVE A COMMENT HERE: