Considering the available descriptive and preliminary mechanistic findings on the PDAC tumor microbiome, the question of its potential diagnostic value and possible implication as a biomarker may arise. One of the main problems with PDAC is most often the late-stage diagnosis as the tumor is often locally advanced or metastasized. This is mostly due to a lack of early-stage symptoms. To date, a reliable screening method for pancreatic cancer is not available in the clinical routine [63]. Studies investigating different site-specific microbiomes, such as the oral and fecal microbiome, point towards a possible application of the microbiome as a diagnostic biomarker in PDAC [49,64].

Indeed, there are numerous publications addressing the microbiome in the oral cavity and its diagnostic potential for PDAC, of which the latest are summarized in Table 2. One of the largest studies was published by Fan et al., which was a population-based nested case-control study on the predictive power of the oral microbiome to assess the risk for pancreatic cancer [65]. Over 730 oral wash samples from two prospective cohort studies were evaluated. The authors found oral pathogens such as Porphyromonas gingivalis to be associated with an increased pancreatic cancer risk. The pitfall of the microbial patterns of the oral cavity, however, is their rather pronounced heterogeneity and low specificity, as they may also be present in other cancer entities [66]. Microbiome studies present contradictory results concerning the microbial composition and differential abundances of these microbes (Table 2). This can be mainly ascribed to the different kinds of sampling methods, e.g., sputum, dorsal tongue, buccal, or gingival swabs. Furthermore, due to different sequencing approaches, i.e., depending on the selected variable (V) region of the 16S rRNA gene, the results significantly vary [67].

Many studies have demonstrated a positive correlation between the pancreas and gut microbiome. For example, Ren et al. reported that the gut microbiome analyzed via stool samples was unique in PDAC and may serve as a non-invasive biomarker for the diagnosis of this disease [70]. Recently, Kartal et al. explored the fecal and salivary microbiota in PDAC patient samples from a Spanish and German case-control study as potential biomarkers; they found 27 fecal species that could be employed to identify PDAC throughout early and late stages with high accuracy. Thus, the authors suggested the fecal microbiome as a feasible early-stage PDAC biomarker, particularly in combination with carbohydrate antigen 19-9 [64]. However, these findings require validation in larger patient cohorts. Only a few months later, Nagata et al. reused the data from Kartal et al. and added their Japanese cohort dataset, which also included oral and gut bacteriophages [84]. Their aim was to further identity oral and gut metagenomic microbial signatures to predict PDAC. The authors found 30 gut and 18 oral species to be significantly associated with PDAC in their newly introduced Japanese cohort, and their metagenomic classifiers were also able to predict PDAC accurately. Consistently with Kartal et al., Nagata et al. found the gut microbiomes of European and Asian patients to present a globally robust and powerful biomarker for identifying PDAC.

Taken together, the orointestinal microbiome might be used as a non-invasive screening tool. However, the translational implication to the clinical setting remains unclear at present. Given the high cost of sequencing, a multiplex PCR or microarray for those identified bacteria might be more feasible. Furthermore, it must be discussed who will be screened, whether it be only high-risk patients or a broader screening population. Further studies with high sample numbers, such as one already completed in the U.S. (NCT03302637), will hopefully provide answers to these questions.

One of the most common sampling techniques in the clinical routine is blood drawing. Bacterial extracellular vesicles (bEV) are nano-sized, lipid membrane-delimited particles that contain different molecules, such as DNA, metabolites, proteins, and lipids. Recently, there is growing evidence that bEVs play an important role in bacteria–bacteria and bacteria–host communication [85,86]. These bEVs can be detected in the host’s blood, urine, bile, and stool. The exploitation of these vesicles for therapeutic and diagnostic purposes is still in its infancy [87]. One recent study revealed the diagnostic value of bEVs for differentiating between benign and malignant tumors [88]. Another Korean study performed a retrospective propensity score matching analysis showing a distinguishable composition of bEVs in blood by 16S rRNA sequencing [89]. Here again, environmental bacteria were detected in peripheral blood, emphasizing the need for thorough decontamination protocols for blood samples as well, in cases where bacterial DNA is found in very low concentrations [85]. Poore et al. demonstrated that microbial plasma profiles in over 10,000 patients, which were different from their respective healthy tissue signature, can predict different cancer types [90]. The authors used whole-genome and whole-transcriptome sequencing studies from TCGA. Moreover, pre-diagnosis blood samples from PDAC patients were subject to oral microbiota antibody measurements in a study by Michaud et al. Indeed, high levels of antibodies against Porphyromonas gingivalis, the pathogen responsible for periodontitis, was correlated with a two-fold increased PDAC risk [48].