Gut microbiome changes associated with chronic pancreatitis and pancreatic cancer: a systematic review and meta-analysis

Background: The study of changes in the microbiome in chronic pancreatitis (CP) and pancreatic ductal adenocarcinoma (PDAC) holds significant potential for developing noninvasive diagnostic tools as well as innovative interventions to alter the progression of diseases. This systematic review and meta-analysis aimed to analyze in detail the taxonomic and functional characteristics of the gut microbiome in patients with CP and PDAC. Methods: Two researchers conducted a systematic search across public databases to gather all published research up to June 2023. Diversity and gut microbiota composition are the main outcomes the authors focus on. Results: This meta-analysis included 14 studies, involving a total of 1511 individuals in the PDAC (n=285), CP (n=342), and control (n=649) groups. Our results show a significant difference in the composition of gut microbiota between PDAC/CP patients compared to healthy controls (HC), as evidenced by a slight decrease in α-diversity, including Shannon (SMD=−0.33; P=0.002 and SMD=−0.59; P<0.001, respectively) and a statistically significant β-diversity (P<0.05). The pooled results showed that at the phylum level, the proportion of Firmicutes was lower in PDAC and CP patients than in HC patients. At the genus level, more than two studies demonstrated that four genera were significantly increased in PDAC patients compared to HC (e.g. Escherichia-Shigella and Veillonella). CP patients had an increase in four genera (e.g. Escherichia-Shigella and Klebsiella) and a decrease in eight genera (e.g. Coprococcus and Bifidobacterium) compared to HC. Functional/metabolomics results from various studies also showed differences between PDAC/CP patients and HC. In addition, this study found no significant differences in gut microbiota between PDAC and CP patients. Conclusions: Current evidence suggests changes in gut microbiota is associated with PDAC/CP, commonly reflected by a reduction in beneficial species and an increase in the pathogenic species. Further studies are needed to confirm these findings and explore therapeutic possibilities.


Introduction
The human microbiota is a leading field of dynamic research in our time, with a predominant focus on the distal gastrointestinal tract, which harbors most of our microbes [1] .The human gut is home to a vast community of 100 trillion microorganisms encompassing more than 1000 distinct resident bacterial species [2] .In the digestive tract, the predominant bacteria are Firmicutes and Bacteroidetes, which account for ~80-90% of the population [3] .Human physiology is significantly impacted by the gut microbiota, as it influences metabolism, regulates the immune system of mucosal tissues, aids in digestion, and maintains intestinal architecture [4] .Dysbiosis of the gut microbiota is linked to various illnesses, including localized gastrointestinal disorders and metabolic, respiratory, hepatic, and neurological disorders [5] .
Intrinsic diseases of the pancreas, pancreatitis, and pancreatic ductal adenocarcinoma (PDAC), cause significant morbidity and mortality in the population [6] .Previously, it was believed that the normal pancreas did not come into direct contact with the intestinal microbiota and thus did not have its own microbiome.The presence of pancreatic microbiota has been demonstrated in a variety of normal and diseased conditions recently [7][8][9] .Despite this, the routes of bacterial entry into the pancreas remain controversial [6,10] .Pancreatic diseases, including acute pancreatitis (AP), chronic pancreatitis (CP), and pancreatic cancer (PC), are now identified as having altered gut microbiotas [4,11] .Recently, a systematic review reported that changes in microbial composition were associated with both AP and CP [12] .It has been reported that 5 years after diagnosis, patients with CP have a nearly eight-fold increased risk of developing PDAC, suggesting that CP is a risk factor for PDAC [13] .Therefore, it is significant to clarify the microbiota imbalance in these pancreatic diseases, especially in patients with CP and PDAC.
This systematic review aimed to examine the taxonomic and functional characteristics of the gut microbiome in individuals with CP and PDAC.By summarizing these findings, we can determine the relationship between alterations in the gut microbiota and worsening pancreatitis or driving cancer progression.The study of changes in the microbiome in CP and PDAC holds significant potential for developing noninvasive diagnostic tools as well as innovative interventions to alter the progression of these diseases.

Method
This study was conducted following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), MOOSE (Meta-Analysis of Observational Studies in Epidemiology), and AMSTAR (Assessing the methodological quality of systematic reviews) Guidelines (Supplementary Table S1, Supplemental Digital Content 1, http://links.lww.com/JS9/C704,Supplementary Table S2, Supplemental Digital Content 2, http://links.lww.com/JS9/C705) [14][15][16] .This study was performed using publicly accessible data sourced exclusively from previously approved ethical studies.The research protocol for this study has been duly registered in the International Registry of PROSPERO (CRD42023484881).

Literature search strategy
An independent and comprehensive search was performed in several public databases, with PubMed, Embase (via OVID), Web of Science, and Cochrane Library by two reviewers.The aim was to identify all original research related to the topic.The following terminologies were used: 'microbiota', 'gut microbiota', 'gastrointestinal microbiome', 'pancreatic cancer', 'pancreatic carcinoma', 'pancreatic', and 'pancreatitis'.Supplementary Table S3 (Supplemental Digital Content 3, http://links.lww.com/JS9/C706) provides additional details regarding the search strategy employed.The last identification date was June 2023.To identify potential research studies which may have been overlooked during the initial database search, further manual searches were conducted within the relevant literature.

Inclusion and exclusion criteria
The studies were selected based on the following criteria: (1) the study compared the gut microbiota of healthy people with people suffering from PDAC or CP; (2) observational studies (such as cohort studies, case-control studies, and cross-sectional studies).Exclusion criteria: (1) the samples collected in the study are not gut-related samples, such as blood, oral samples, tissue samples, etc.; (2) animal studies were excluded; (3) the study was presented as a conference abstract, case study, or narrative review; (4) the study was not authored in the English language.

Data extraction and quality assessment
Two reviewers independently screened the titles and abstracts of the retrieved studies.In case of any discrepancies, they engaged in collaborative discussions to reach a consensus, seeking guidance from a third experienced reviewer whenever necessary.Following are the details extracted: author name, year of publication, country, type of study design, characteristics of participants, type of samples, genetic analysis, and result of the gut microbiome.To assess the methodological quality of the observational studies, the Newcastle-Ottawa Scale (NOS) was applied [17] .A score of 7-9 indicated superior quality.

Statistical analysis
To address differences in gut microbiota, we conducted a literature review and reported partial meta-analysis results at the phylum level.A descriptive synthesis was conducted since microbiota proportions were assessed differently, as well as the quality of included studies was inadequate.Meta-analysis were carried out using Comprehensive Meta-Analysis 3.0.The Shannon Index, Simpson Index, Evenness, and Richness of bacteria were calculated as standardized mean differences (SMDs) between the PDAC, CP, and control groups [18] .We used the inverse variance method to pool odds ratios (ORs) with corresponding 95% CIs to determine the composition of gut microbiota at the phylum level.For continuous variables where median and interquartile range were provided by studies instead of mean and SD, we utilized Wan et al.'s [19] method to calculate them.Statistical significance was set at P < 0.05.Heterogeneity was measured using I 2 statistics for all studies; fixed-effects models were applied when I 2 < 50%, while random-effects models were applied when I 2 > 50%.In cases where there was significant heterogeneity among included studies in the meta-analysis (n > 10), funnel plots were employed to observe publication bias or other biases that may have affected results obtained from these sources.Finally, sensitivity analyses involved excluding one study

HIGHLIGHTS
• This is the first systematic review and meta-analysis to focus specifically on the gut microbiota of patients with chronic pancreatitis (CP) and pancreatic cancer (PC).• Compared to the healthy controls, significant differences in gut microbial composition between PC/CP patients, as evidenced by a somewhat reduced α-diversity and statistically significant β-diversity.
• At the phylum level, the proportion of Firmicutes was less abundant in patients with PC and CP than in healthy controls.At the genus level, changes in gut microbiota associated with PC/CP, are commonly reflected by a reduction in beneficial species and an increase in the pathogenic species.
after another sequentially until final results or heterogeneity changed significantly enough for us to take note thereof during our analysis process.

Result
Study selection A total of 3966 research studies were identified as a result of the systematic literature review.After removing duplicate records, we assessed a total of 1504 records based on their title and abstracts.Subsequently, the search yielded a selection of 50 articles which may have fulfilled all the specified criteria for inclusion.Out of these, 36 studies were excluded due to various reasons such as utilization of animal models instead of human subjects, absence of both healthy individuals and CP/PDAC patients in the study groups, usage of saliva or blood samples rather than gut-related samples, and unavailability of data.The final meta-analysis included 14 studies that adhered strictly to the inclusion and exclusion criteria [11,[20][21][22][23][24][25][26][27][28][29][30][31][32] .The PRISMA flow chart illustrating this meta-analysis can be observed in Figure 1.

Study characteristics and quality assessment
Four cohort studies, eight case-control studies, and two crosssectional studies were included.
Of the nine studies that reported CP versus HC groups, six had statistically significant differences in the intestinal microbiota composition between them, and one had no significant differences.The remaining two items had no significant difference at PCoA of weighted UniFrac distances but had statistical differences at PCoA of unweighted UniFrac distances.
Three studies reported the comparison of β-diversity between the CP group and PDAC group, and all three studies showed statistically significant differences in the composition of gut microbial community between PDAC and CP.

Summary and meta-analysis of gut microbiota composition
All 14 studies reported significant microbiota changes in patients with PDAC/CP.We summarized the classification of microorganisms that differed significantly at various levels (e.g.phylum, genus, and species level) in PDAC/CP patients compared to HC (Table 2).
Compared to HC, the gut microbiota composition in PDAC patients is different.At the phylum level, more than two studies presented an increase in the proportion of Proteobacteria and Fusobacterium phylum in patients with PDAC.At the genus level, 22 genera were identified in five studies.The following results were found in more than two studies with the same trend: Actinomyces, Bifidobacterium, Veillonella, and Escherichia-Shigella were increased in patients with PDAC compared to HC.At the phylum level, we conducted a meta-analysis of gut microbiota, but the pooled results showed that compared to the HC group, the proportion of Firmicutes in PDAC patients was significantly reduced (OR = 0.63; 95% CI: 0.42-0.95;P = 0.029; I 2 = 0%), but there was no significant difference in other phyla (Actinobacteria, Bacteroidetes, Fusobacteria, and Proteobacteria) (P > 0.05) (Fig. 2).There were also differences in the gut microbiota of patients in the CP group compared to the HC group.At the phylum level, more than two studies reported an increase in the proportion of Proteobacteria and a decrease in the proportion of Bacteroidetes and Actinobacteria in patients with CP.At the genus level, 52 genera were identified in seven studies.The following results were presented in more than two studies with the same trend: Escherichia-Shigella, Enterococcus, Klebsiella, and Streptococcus were increased in CP patients, while Faecalibacterium, Bifidobacterium, Coprococcus, Subdoligranulum, Parasutterella, Paraprevotella, Fusicatenibacter, and Roseburia and were decreased compared to HC.At the phylum level, we conducted a meta-analysis, and the pooled results showed a significantly lower proportion of Firmicutes in patients with CP compared to the HC group (OR = 0.68; 95% CI: 0.46-0.99;P = 0.049; I 2 = 0%), but there was no significant difference in other phyla (Actinobacteria, Bacteroidetes, and Proteobacteria) (P > 0.05) (Fig. 3).
In addition, two studies reported differences in gut microbiota at the phylum level between the PDAC and CP groups, and our pooled results showed no significant difference in the proportions of Bacteroidetes, Firmicutes, and Proteobacteria between the CP and PDAC groups (P > 0.05) (Fig. 4).Likewise, at the genus level, no more than two studies reported significant changes in the microbiota with the same trend.

Summary of functional/metabolomic change
Compared to the HC group, the PDAC group had a significant increase in lipopolysaccharide (LPS) synthesis [20,23] , phosphotransferase systems, ATP-binding cassette transporters, and terpenoid backbone biosynthesis [11] , activation of inflammatory pathways, and decreased cell motility [20] .In terms of metabolic pathways, the PDAC group had a higher potential for degrading fatty acids, a significantly lower metabolic capacity for the synthesis of short-chain fatty acids (SCFA), a higher potential for putrescine and spermidine transport [23] , and a significant decrease in propionic acid and deoxycholic acid [21] , amino acid, secondary metabolite biosynthesis, and in porphyrin and chlorophyll metabolism [11] .Similarly, the LPS synthesis pathway [26,28] and bacterial invasion pathway [28] were enriched in the CP group compared to the HC, increasing plasma endotoxin levels [26] .Additionally, the phosphotransferase system of the CP group was enriched while ribosomal activity was reduced [29] .In terms of metabolism pathways, chlorophyll metabolism, and porphyrin, sucrose and starch metabolism are depleted [29] , arginine and proline metabolism, as well as glycolysis and gluconeogenesis pathways, are severely depleted [28] .

Publication bias and sensitivity analysis
A 0% I 2 value indicates minimal heterogeneity among studies of gut microbiota composition between PDAC/CP and HC groups.The heterogeneity of the combined results of the meta-analysis is shown in detail in Supplementary Table S6 (Supplemental Digital Content 7, http://links.lww.com/JS9/C710).Funnel charts should not be used as the number of included studies for each indicator did not exceed ten.To assess if a single research would affect the overall results, we used sensitivity analysis.After each study was deleted individually, the analysis was largely stable (Supplementary Fig. S2, Supplemental Digital Content 8, http://links.lww.com/JS9/C711).

Discussion
In the past, several reviews have explored the role of the human microbiome in pancreatic disease [12,33,34] , but to our knowledge, this is the first systematic review to focus specifically on the gut microbiota of patients with CP and PDAC.In this meta-analysis of 14 studies, we systematically assessed the correlation between gut microbiota and PDAC/CP.First, we found differences in the composition of gut microbiota between PDAC/CP patients compared to the HC, as evidenced by a somewhat reduced α-diversity and statistically significant β-diversity.Second, at the phylum level, the proportion of Firmicutes was less abundant in PDAC and CP patients than in HC.Third, at the genus level, Actinomyces, Bifidobacterium, Veillonella, and Escherichia-Shigella were increased in PDAC patients compared to HC. Streptococcus, Klebsiella, Enterococcus, and Escherichia-Shigella were increased in CP patients, while Faecalibacterium, Bifidobacterium, Coprococcus, Subdoligranulum, Fusicatenibacter, Roseburia, Parasutterella, and Paraprevotella were decreased compared to HC.Finally, functional/metabolomics results from various studies also showed differences between patients with PDAC/CP and the HC.Additionally, our study finds no significant differences in intestinal microbiota between PDAC patients and CP patients.
While the diversity metrics themselves may not have enough resolution to serve as diagnostic or prognostic markers, analyzing diversity is a convincing tool for broadly evaluating the composition of the gut microbiota.It offers a less biased approach compared to culture-dependent or limited targeted assay methods [12,35] .Regarding diversity assessment, to our knowledge, none of the previous studies have performed meta-analyses of α and β diversity assessments in PDAC and CP patients.Our meta-analysis revealed that both the PDAC group and the CP group had lower Shannon index values compared to the HC group, with no significant difference between the PDAC and CP groups.The Shannon index primarily considers two factors: the number of species as operational taxonomic units and their relative abundance inequality [36,37] .However, due to the fact that each study includes a different number of species, the Shannon index does not fully assess inequality.One study showed that long-surviving PDAC patients have higher α-diversity in their tumor microbiome, and it has been demonstrated through fecal microbiome transplantation (FMT) experiments that the composition of the microbiome in PDAC that cross-talks to the gut microbiome, influences the host immune response and natural history of the disease [38] .Regarding β-diversity analysis, we were unable to conduct a meta-analysis due to insufficient data; therefore we can only describe the results of individual studies in detail.We found that most studies reported significant differences between the PDAC/CP and HC groups.Also, interestingly, there were differences between the PDAC and CP groups.It has been shown in multiple studies that CP disease severity is linked to significant changes in the composition of gut microbiome and reduced relative abundance of commensal organisms [26,28] .Regardless of the underlying cause of CP, similar discoveries have been documented [23,39] , suggesting that changes in the microbial composition may be an effect rather than a cause of CP [12] .Therefore, there is a clear need to further investigate the differences in microbial species diversity among PDAC, CP, and HC patients.
The abundance of Firmicutes was decreased in the gut of PDAC patients and CP compared to HC patients, based on pooled results.Firmicutes consist of a large number of core bacteria with more diverse functions.From a metabolic point of view, these bacteria play a concernful role in host and gut homeostasis through the production of SCFA, which normalizes intestinal permeability.An increase or decrease in the Firmicutes/ Bacteroidetes ratio is considered an ecological imbalance [40] .The former is often associated with metabolic disorders and obesity.The latter is associated with depression, inflammatory bowel disease, etc., which may be related to the immunoinflammatory response induced by the reduced production of SCFA, especially butyric acid, histamine, LPS accumulation, etc. [41] .Faecalibacterium and Coprococcu are among the most abundant commensal bacteria in the intestine and are members of the Firmicutes phylum.In addition, they are an important source of butyric acid production in their gut microbiota.Multiple studies have consistently presented a clear reduction in butyric-producing bacteria in CP patients, such as Faecalibacterium, Coprococcu, Roseburia, and Fusicatenibacter.The decrease of butyric acid, butyrate, and other strains reduces the levels of tight junction protein and SCFA, increases the permeability of human intestinal mucosa and the dysfunction of the intestinal mucosal barrier, resulting in the release of intestinal pathogens into the blood through the damaged intestinal mucosa, thereby promoting the occurrence of pancreatitis [42] .At present, we can only speculate that there may be a potential association with the occurrence of pancreatic inflammation, but the specific causal relationship needs to be further demonstrated.
Bifidobacteria are anaerobic bacteria and are considered to be an important beneficial microbe in the gut.Bifidobacterium is often used in combination with Lactobacillus to treat chronic diarrhea and constipation and to assist in the treatment of endotoxemia caused by intestinal microbiota imbalance.One study reported a lower abundance of Bifidobacterium in the intestinal of PDAC patients compared to HC [22] , whereas one studies reported a higher abundance of Bifidobacterium in PDAC patients (duodenal fluid samples) [24] .Correspondingly, studies have reported that repopulation of the germ-free KC genetic mouse model of pancreatic tumors, with Bifidobacterium pseudopodium-accelerated pancreatic tumorigenesis and can be detected in the pancreas of treated mice [9] .However, many recent studies have found that Bifidobacterium strains enhance the efficacy of anti-PD-L1 treatment and synergistically reduce tumor burden by inducing antitumor host immune response [43][44][45] .These studies and findings contribute to the application of strategies to modulate the intestinal microbiota to cancer immunotherapy in clinical practice, providing evidence for the treatment of PDAC with live microbial products.Two studies reported lower abundances of Bifidobacterium in the intestinal of CP patients compared with HC [29,32] .In a recent study, fecal Bifidobacteria abundance was found to be inversely correlated with systemic inflammation in patients with pancreatitis.It has also been found that Bifidobacteria and its metabolite lactate prevent pancreatitis in mice by alleviating macrophage-associated pancreatic and systemic inflammation through the combined action of TLR4/MyD88 and NLRP3/Caspase1 [46] .Several studies have reported significantly higher abundances of Escherichia-Shigella in the intestines of both PDAC and CP patients compared to HC [20,23,30,31] .Escherichia-Shigella is the most common pathogen of bacterial dysentery in humans, producing endotoxins and exotoxins that can damage the intestinal mucosa.Bacterial invasion of epithelial cells in pancreatitis has been reported, highly correlated with the abundance of Escherichia-Shigella [47] .In addition, studies have found significant changes in the gut microbiota of mice with or without pancreatitis after NLRP3 knockout, suggesting that NLRP3 deficiency can counteract pancreatitis-induced microbial interference [48] .Several studies have reported an increase in Klebsiella in the intestinal of CP patients compared to HC [27,29] .Klebsiella is significantly elevated in pancreatic cysts or pancreatitis fluid, suggesting that the intestine is likely to be the pathway for Klebsiella to invade the pancreas [49,50] .
Establishing a distinct correlation between alterations in the gut microbiota and pancreatitis/PDAC holds significant significance, given its potential to offer valuable avenues for clinical intervention.However, based on the current evidence, we cannot investigate whether dysbiosis is pathogenic or reactive.Therefore, it is also indispensable to explore the pathogenic mechanism, for example, intestinal barrier dysfunction, bacterial translocation, microcirculation dysfunction and autodigestion may also be related to pancreatic inflammation [51,52] .Further studies are ongoing, for example, in PROCEED, immunoassay results are reported for serum samples collected from large pancreatitis cohorts, identifying immune markers that could act as potential biomarkers to distinguish different pancreatitis diseases [53] .However, the results of whether CP and diabetes alter the microbiome and lead to PDAC have not been published [54] .A recent two-sample Mendelian Randomization (MR) analysis to investigate the causal relationship effect between intestinal microbiota and PDAC showed that Senegalimassilia showed protective effects against PDAC, while Streptococcus, Odoribacter, Ruminiclostridium 9, and Ruminococcaceae (UCG011) were identified as causative factors for PDAC [55] .Another MR study involving 18 340 participants in a genome-wide association study (GWAS) explored the causal effect between intestinal microbiota and four types of pancreatitis (acute and CP and alcohol-induced acute and CP), revealing correlations between pancreatitis and 30 different types of intestinal microbiota [56] .Identifying these potential associations will be important for preventing and treating pancreatic disease, as well as determining how to apply strategies that alter the gut microbiota, for example, antibiotics, probiotics, dietary changes, and FMT.
This systematic review and meta-analysis has several limitations.First, the selected studies had a wide range of participants.Several of the included studies involved subjects with specific types of pancreatitis, including autoimmune pancreatitis, alcoholic pancreatitis, childhood pancreatitis, and pancreatitis with diabetes [23,26,27,29] .Age, alcohol, metabolism, and immunity may affect the distribution of microbiota [57] .Second, different methods of genetic analysis were used in the included studies, including sequencing methods, regional specifications, pipeline analyses, and database.It is possible that the results were influenced by inconsistencies in methodology, specifically in specimen handling, DNA extraction protocols, and sequencing methods.Third, although the included studies were conducted on a global scale, there was not enough information to allow us to perform a subgroup analysis comparing different geographical zones.Due to dietary differences between countries, there are conspicuous differences in the composition of nutrients among Asians, Americans, and Europeans, which will have a considerable influence on the composition of intestinal microbiota.Fourth, though our study found correlations between the gut microbiome and PDAC/CP, the observational nature of the study makes it difficult to determine a causal relationship between the gut microbiome and PC.Fifth, the number of studies examining the relationship between PDAC and CP is small, so it is not possible to produce more reliable results.Sixth, several studies presented taxonomic data at different taxonomic levels as well as reported species-level data obtained using Metagenomics sequencing.

Conclusion
In conclusion, current evidence suggests changes in gut microbiota associated with PDAC/CP, commonly reflected by a reduction in beneficial species and an increase in the pathogenic species.The difference between PDAC and CP has not been found.Functional/metabolomic results from various studies also show differences between patients with pancreatic disease and HC.Detection of differences in microbial composition associated with CP and PDAC could serve as a biomarker and be a diagnostic tool.However, further research is needed to elucidate the potential role of the gut microbiome in pancreatic disease, which may provide evidence for biomarkers and microbial therapies.

Ethical approval
Not applicable (this paper was provided based on research in global databases).

Consent
Not applicable (this paper was provided based on research in global databases).

Figure 2 .
Figure 2. Forest plot of meta-analysis of the composition of gut microbiota at phylum level (pancreatic cancer versus healthy individuals).

Figure 3 .
Figure 3. Forest plot of meta-analysis of the composition of gut microbiota at phylum level (chronic pancreatitis versus healthy individuals).

Figure 4 .
Figure 4. Forest plot of meta-analysis of the composition of gut microbiota at the phylum level (chronic pancreatitis versus pancreatic cancer).
Three studies reported relevant data in the PDAC and CP groups, four studies reported data only in the PDAC group, and seven studies reported data only in the

Table 1
Characteristics of all the studies included in the meta-analysis.

Table 2
Genetic analysis and microbial change (patient vs. control) in the included studies.

Table 3
The meta-analyses of alpha-diversity outcomes of the included studies.