Detection of MAP in Crohn’s Disease Patients

Earlier systematic reviews have concluded that MAP detection is significantly enriched in Crohn's disease patients (Feller et al., 2007; Abubakar et al., 2008; Waddell et al., 2008).

Isolation in pure culture from diseased tis­sue is classically the gold standard for demon­strating the presence of a bacterium and serves as the first step in determining if the bacterium is associated with a particular disease (if found to be more common in diseased than in control subjects). However, this method comes with a set of challenges, some inherent to MAP micro­biology and some inherent to research involving human subjects. MAP is among the slowest- growing culturable mycobacteria, and proper decontamination is difficult to achieve in tissue samples containing many other competing bac­teria. Yet, other chapters of this book provide details on how to isolate this same slow-growing bacterium from manure, demonstrating techni­cal feasibility. Complicating factors in human studies include the nature of the sample and the status of the patient. In Crohn’s patients who are clinically stable, one can only obtain mucosal bi­opsies or stool samples. If the organisms reside beneath the mucosa (e.g. the muscularis, the mesenteric lymph nodes), then the chance of de­tecting them is greatly reduced. In most Crohn’s patients who have progressed to the point of re­quiring a surgical excision, there has been a trial of pre-operative antibiotics to avoid surgery (e.g.

Overall, recent reports have detected MAP by culture in a small percentage of Crohn’s dis­ease patients. One study cultured MAP from 2/18 (11.1%) Crohn’s patients, although none could be cultured from ulcerative colitis or non- IBD controls (Zamani et al., 2017). Banche et al. (2015) cultured MAP from 13/76 (17.1%) sam­ples from patients with Crohn’s disease com­pared with 2/44 (4.5%) controls. Another study did not culture MAP from Crohn’s disease pa­tients, however only five patients were included in the study (Carvalho et al., 2016). In a larger study including 75 individuals recently diag­nosed with Crohn’s disease, two (2.7%) were MAP positive by culture compared with 2/135 (1.5%) controls (Ricanek et al., 2010).

Due to the difficulties with MAP culture, most studies combine culture with molecular methods such as IS900 PCR to detect MAP. However, the presence of bacterial DNA does not equate with actively growing bacteria, as exemplified by the use of ancient DNA in an­thropology. The majority of molecular method­based studies have reported a higher frequency of MAP in Crohn’s disease patients, but they are also subject to considerable heterogene­ity. Indeed, many recent studies have reported a significant association with Crohn’s disease (Singh et al., 2011b; Banche et al., 2015; Khan et al., 2016; Zamani et al., 2017); however, PCR detection rates continued to range from 0-100% across studies (Parrish et al., 2009; Sasikala et al., 2009; Mendoza et al., 2010). A 2015 systematic review indicated that a sig­nificant amount of heterogeneity could be ex­plained by choice of primers and the sampling frame (Waddell et al., 2015). Even within the rubric of PCR-based detection of IS900, there are technical variants, such as the nested IS900 PCR assay.

Another method used to detect MAP in tis­sue is staining. Staining for acid-fast bacilli and looking under oil-immersion microscopy was described by Jeyanathan et al. (2007), who ap­plied methods previously validated on tissue from mice experimentally infected with MAP (positive control) and M. tuberculosis (negative control) (Jeyanathan et al., 2006). Banche et al. (2015) reported non-significant differences in detection where 3/76 (3.95%) Crohn’s dis­ease samples were positive by acid-fast staining compared with 0/44 (0%) non-IBD controls. Zarei-Kordshouli et al. (2019) detected acid-fast bacteria in 2/30 (7%) Crohn’s disease subjects compared with 0/30 (0%) non-IBD controls and 27/30 (90%) caprine paratuberculosis sam­ples. Studies using immunohistochemistry also failed to detect differences between samples from Crohn’s disease and controls (Magin et al., 2013; Khan et al., 2016). However, MAP infection in humans is often observed to be paucibacillary, which can be more difficult to visualize using acid-fast staining. In addition, some studies have suggested that this lack of visualization is due to MAP entering a cell wall-deficient spheroplast form in human tissue, which will not be detect­ed by acid-fast staining (Chiodini et al., 1986). Given that cell wall-deficient forms of MAP have not to our knowledge been described in the natural bovine and ovine hosts, this would be a unique situation in clinical microbiology where a zoonotic agent undergoes a morphologic tran­sition during the human spillover infection. An alternative possibility is that MAP does undergo a transformation into a cell wall-deficient form in its natural hosts, but this has not been report­ed to date.

There is some evidence to suggest intra­study variability in results due to changes in MAP status in individuals. In 2009, Kirkwood et al. reported significant MAP association in biopsies and peripheral blood mononuclear cells (PBMCs) of children with Crohn's dis­ease by IS900 PCR and culture compared with non-IBD controls (Kirkwood et al., 2009). The children participating in this study were then followed up for up to 4 years. MAP was found to persist only within a subset of patients and the authors suggested that infection status at disease onset may be an important factor in predicting infection persistence (Wagner et al.,

2011). Overall, the recent data suggest that, although a positive association between MAP and Crohn's disease has been consistently re­ported by systematic reviews, heterogeneity between and within studies continues to be re­ported. Future studies applying multiple exter­nally validated and standardized methods may allow for more robust cross-study comparison.

3.5