Is Bartonella a Cause of Primary Sclerosing Cholangitis? A Case Study

Round 1

Reviewer 1 Report

In this study, the author reports a case showing the association between Bartonella infection and PSC. The topic is interesting but some data are missing. The author claims that Bartonella infection causes autoimmune responses leading to PSC in conclusion. However, the similar condition caused by autoimmunity is called primary biliary cholangitis (PBC), and it is known that alphaproteobacterial infection induces autoimmune diseases such as PBC (Mohammed and Mattner, 2009). The patient in this study might be PBC, not PSC, driven by Bartonella infection. This study lacks evidence that the patient was indeed PSC but not PBC showing AMA levels etc. In addition, the patient has a history of Lyme disease. Lyme disease is caused by infection and inflammatory responses can cause biliary damage leading to cholestasis (PSC). PSC is progressive but slow, so it is possible that the patient had Lyme disease followed by biliary inflammation, and that lead to PSC-like symptoms later regardless of Bartonella infection. Overall, this case reports lacks evidence to prove whether the patient was PSC or PBC as well as whether symptoms and liver conditions were caused by Bartonella infection or Lyme disease.

Author Response

Reviewer 1

Q: Does this patient have primary biliary cirrhosis (PBC) rather that primary sclerosing cholangitis (PSC)?

A: I agree with this reviewer that PBC needs to be considered as an alternative diagnosis to PSC. Unfortunately, an AMA titer was not done, and the patient was transferred out of my care once the diagnosis was made. When considering the differential diagnosis, most people who get PBC are women from 30 to 60 years of age. While inflammatory bowel disease has occurred in some patients with PBC, it is an uncommon phenomenon (Shizuma, 2014). Furthermore, PBC primarily involves cholangitis in the intrahepatic ducts, while PSC involves both the intrahepatic and extrahepatic ducts. This patient presented with issues involving only the extrahepatic ducts. I believe that the findings support the diagnosis of PSC.

Q: Does this patient have Borrelia burgdorferi as a cause of biliary inflammation since this organism has been found in patients with inflammatory bowel disease?

A: The reviewer raises an intriguing question regarding whether the causative organism of Lyme disease is responsible for both this patient’s inflammatory bowel disease and cholangitis. It has been reported that both B. burgdorferi as well as Bartonella can cause inflammatory bowel disease, and they can both stimulate autoimmune reactivity. As well, vancomycin is an effective drug against both organisms. While B burgdorferi has been associated with an elevation in liver enzymes and hepatitis (usually in acute infections) and occasionally bilirubin, there are no reports of cholestasis or cholangitis in the medical literature, and the patient did undergo a prolonged treatment of his Lyme disease with amoxicillin, which would likely treat Lyme disease but not adequately treat bartonellosis. There is more documentation that Bartonella is a cause of inflammatory bowel disease than B. burgdorferi, and there is documentation that Bartonella can cause cholangitis (Kordick, 1999). As well, the abdominal ultrasound suggested the presence of granulomas; Bartonella is a documented cause of granulomas in both the liver and spleen, but B. burgdorferi is not. It is intriguing to consider the possibility that B. burgdorferi could cause inflammatory bowel disease and trigger an autoimmune response manifesting as cholangitis.

Reviewer 2 Report

 

Overall comments:

 

This is a very interesting case report and beautifully written. The author presents an intriguing case of classical PSC/IBD preceded by a known Bartonella infection, which all resolved with vancomycin treatment.

 

This link has not been postulated in humans before and is an interesting theory, though very much speculation/association at this stage.

 

It would be useful to have representative images of the MRCPs pre- and post- vancomycin as well as endoscopic pictures of the colon pre- and post. It is important to know that his colitis actually objectively improved/normalised. Also, a little more detail about the phenotype of his colitis would be useful and the time frame of his colitis symptoms and liver biochemistry derangement. If these two were fairly acute (i.e. weeks rather than months), then I would suspect an infective colitis with a secondary sclerosing cholangitis, rather than true IBD and PSC. Finally, a bit more detail on how long he was treated with the vancomycin, the dose of mesalamine (as this could have been the thing to heal the colitis), and the duration to resolution of symptoms/Ix would also be helpful.

 

Overall, it is an original and interesting addition to the literature on PSC, and, depending on the above, merits further future investigation.

 

 

Minor points:

Reference range abbreviation should be capitalised  = RR Ferritin reference range differs from units given for case Calprotectin should read “faecal calprotectin” I would comment that liver enzymes were raised in a “cholestatic” pattern, and include GGT Indirect Immunofluorescence assay is usually abbreviated as “IIF” Colonoscopy – what was the pattern of the colitis? Pancolitis? (i.e. a pattern consistent with PSC-IBD)? I would break up some of the case presentation results into different paragraphs rather than one very long paragraph to make it more readable. Or possibly even if with subheadings “the prevalence of non-HLA genome-wide single nucleotide polymorphisms”.. which are also related to immune functions Ursodiol – use generic name UDCA enzyme linked immunoassay (ELI) should be “enzyme linked immunosorbent assay – ELISA” An important oral vanc study in PSC to mention is Rahimpour et al J Gastroint Liv Dis 2016. Should probably refer to the recent case series from Dao A et al (Inflamm Bowel Dis 2019) and Tan LZ et al (Gut 2019) showing improvement in colitis in patients with PSC/IBD A comment on the fact these previous vanc studies are small, predominantly in children rather than adults (lack of beneficial data in adults), and based on ALP changes, which have their own limitations. This may be the first study to show normalisation of MRCP with vancomycin.  Is Bartonella known to be susceptible to vancomycin? How do we know that his Bartonella was eradicated and therefore implicated in the pathogenesis of his illness? If there is no follow up Bartonella testing, this limitation should be acknowledged.

Author Response

Reviewer 2

Q: It would be useful to have representative images of MRCPs and endoscopic pictures of the colon as well as the phenotype of his colitis.

A: I agree 100% with the reviewer that the above information would provide excellent documentation in this patient’s case history. Fortunately, I have been able to procure these images and have added them to the article. I mistakenly wrote that the second MRCP was normal, but after reviewing the actual film it had improved but not normalized.

Q: How long was he treated with the vancomycin, the dose of mesalamine, and the duration to resolution of symptoms?

A: I have added this information to the article. It is noteworthy that the symptom exacerbation and elevated LFTs occurred months after the onset of symptoms, making infective colitis less likely.

Q: Minor points regarding abbreviations, etc.

A: I thank the reviewer for his/her suggestions and I have made appropriate amendments. I have included results of antibody testing. I have researched the appropriate acronym for indirect immunofluorescent antibody assay and it appears to be IFA; when using the term indirect immunoflouresence the acronym is IIF.

I agree with the reviewer re breaking up the discussion directly relevant to the patient into two paragraphs and I have done so.

I have included the recent references regarding treatment of PSC with vancomycin suggested by reviewer 2.

Bartonella is indeed susceptible to vancomycin, and this is documented in the article (reference #59—Sobraques, 1999). We do not know if Bartonella was eradicated, and a negative serology or PCR would not suffice to prove absence of this bacteria. But suppression of this microbe would suffice to suppress local inflammation (i.e. colitis) as well as autoimmune reactivity (cholitis). The fact that many patients with UC and PSC respond to oral vancomycin but relapse when the drug is stopped is consistent with inactivation but not eradication of Bartonella.

Round 2

Reviewer 1 Report

No further comments.

Reviewer 2 Report

They have gone with most of my suggestions, and it reads better than before. Again, to mention, this is an interesting and original case report.