Secondary Logo

Journal Logo

Should Patients With Bartonella Neuroretinitis Receive Treatment?

Bhatti, M. Tariq MD; Lee, Michael S. MD

Section Editor(s): Lee, Andrew G. MD; Van Stavern, Gregory MD

doi: 10.1097/WNO.0000000000000166
Point Counter-Point

Departments of Ophthalmology and Neurology (MTB), Duke University Eye Center, Durham, North Carolina; and Departments of Ophthalmology and Visual Neurosciences, Neurology, and Neurosurgery (MSL), University of Minnesota, Minneapolis.

Address correspondence to Michael S. Lee, MD, 420 Delaware Street SE, MMC 493, Minneapolis, MN 55455; E-mail:

An unrestricted grant to the Duke Eye Center Department of Ophthalmology from Research to Prevent Blindness, Inc (New York, NY).

The authors report no conflicts of interest.

Neuroretinitis refers to focal inflammation of the optic nerve and retina and is typically characterized clinically by painless loss of vision, optic disc swelling, and macular exudates, usually in a star-like pattern. Although a wide variety of infectious and noninfectious diseases are associated with neuroretinitis, Bartonella is the most common infectious etiology. However, whether patients with Bartonella neuroretinitis should be treated remains controversial. Two experts debate this topic.

Back to Top | Article Outline

Pro: Patients With Bartonella Neuroretinitis Should Receive Treatment: Michael S. Lee, MD

Bartonella-related neuroretinitis likely represents an intraocular inflammatory or immune response to systemic exposure to Bartonella henselae or Bartonella quintana rather than true intraocular infection. Among immunocompetent individuals, visual loss is typically self-limited, and the visual prognosis is often excellent without treatment (1). Therefore, even though I believe that most patients with self-limited Bartonella neuroretinitis may not require treatment, I think that there is a compelling rationale for offering treatment to certain patients.

Among patients with systemic Bartonellosis, studies have shown that antimicrobial therapy shortens the symptomatic illness. In many instances, systemic symptoms of fever and lymphadenopathy can accompany Bartonella neuroretinitis, and this may indicate need for therapy (1). B. henselae has shown susceptibility in vitro to rifampin, ciprofloxacin, gentamicin, azithromycin, clarithromycin, and trimethoprim–sulfamethoxazole (2). Bass et al (3) randomized patients with B. henselae lymphadenitis to receive either azithromycin or placebo. At 1 month follow-up, 50% of the azithromycin group achieved reduction of 80% of lymph node volume compared with only 7% in the placebo group. Although I am unaware of any randomized controlled clinical trials for the treatment of Bartonella neuroretinitis, in a retrospective review, Reed et al (4) demonstrated quicker visual recovery and shortened disease course with antibiotic use compared with historical controls. A combination of doxycycline and rifampin (4) or ciprofloxacin (5) has been reportedly beneficial for immunocompetent patients with Bartonella neuroretinitis. However, immunocompromised patients with Bartonella neuroretinitis tend to experience worse visual outcomes. This includes individuals with acquired immunodeficiency syndrome, transplant patients and those with HIV/AIDS. In this patient group, erythromycin or doxycycline can significantly mitigate visual loss (6). Although doxycycline is more effective against Bartonella, erythromycin is preferred for children under 8 years because of tooth discoloration with doxycycline use.

Neuroretinitis represents a descriptive term for optic disc edema with a macular star, and, although B. henselae and Quintana are the most common source, one cannot assume that all cases result from these organisms (7). Toxoplasmosis (8), syphilis (9), Lyme (10), tuberculosis (11), sarcoidosis (12), inflammatory bowel disease (13), polyarteritis nodosa (14), and Rocky Mountain spotted fever (15) among others have been associated with neuroretinitis. At the time of neuroretinitis diagnosis, the underlying cause is not always readily apparent. The physician should inquire about possible risk factors such as cat exposure, potential exposure to ticks or tuberculosis, or sexual contact. Other clues include a positive review of systems for fever, myalgias, headache, productive cough, or skin rash.

In the majority of cases, the exact cause of neuroretinitis is not established by the time the patient leaves the office. Therefore, one might consider treating with antibiotics until the laboratory workup is complete. Many results in the differential diagnosis can return within a day; however, Quantiferon gold can take 2–4 days, and the laboratory results for B. henselae, B. quintana, toxoplasmosis, and Rocky Mountain spotted fever often take several days or more to return depending on the proximity of the testing laboratory. In some cases, acute Bartonella infection shows normal IgM titers. After 4–6 weeks, IgG convalescence titers may rise 4-fold, but this may delay a definitive diagnosis of Bartonella neuroretinitis.

Therefore, I believe that it is reasonable to begin presumptive treatment for neuroretinitis in the following situations: severe visual loss; bilateral involvement; presence of systemic symptoms; immunocompromised state; progressive, protracted clinical course. Bilateral, simultaneous Bartonella neuroretinitis can occur, but more likely represents another infectious or inflammatory etiology (after ruling out malignant hypertension and intracranial hypertension). The final diagnosis of Bartonella neuroretinitis is often made after the acute phase of the disease because of the delay in titer availability. To prevent a delay in treatment for other causes of neuroretinitis, while laboratory testing is in progress, the clinician should consider treatment with azithromycin, which can treat to a reasonable extent toxoplasmosis (16), Lyme (17), and Bartonella (3). I believe that immunocompromised patients with neuroretinitis should receive doxycycline or erythromycin, before the laboratory testing returns (5). In the absence of tuberculosis exposure, I would hold treatment for this before laboratory confirmation. Finally, patients with progressive, protracted, or recurrent neuroretinitis should receive an autoimmune evaluation, and the practitioner may want to initiate systemic corticosteroids as an adjunct to azithromycin.

Back to Top | Article Outline

Con: Patients With Bartonella Neuroretinitis Should Not Receive Treatment: M. Tariq Bhatti, MD

“As to diseases, make a habit of 2 things—to help or at least to do no harm” Hippocrates, Epidemics, Bk. I, Sect. XI.

Any intervention, whether it is the institution of a medication, insertion of a medical device or performing an invasive surgical procedure should always be preceded by a risk-benefit analysis. This holds true for a case of Bartonella neuroretinitis as well. Several important questions that should be addressed in the efficacy and safety of Bartonella neuroretinitis treatment are:

  1. Is B. henselae the cause of neuroretinitis? If so what is the science of disease?
  2. What is the natural history of Bartonella neuroretinitis?
  3. What is the evidence to support treatment and how strong is that evidence?
  4. What are the risks of treatment?

The following discussion will focus on B. henselae infection in an immunocompetent patient. Immunocompromised patients infected with B. henselae present with clinical manifestations and demonstrate disease behavior that is significantly different from patients with a healthy functioning immune system (18).

Back to Top | Article Outline

Is B. henselae the Cause of Neuroretinitis? If So, What Is the Science of Disease?

There is little doubt that B. henselae, an aerobic, oxidase-negative, Gram-negative bacillus, is an arthropod/zoonotic transmitted infectious agent that can result in cat scratch disease (CSD); a self-limited systemic illness associated with fever and lymphadenopathy (19). Aside from the ocular complications of neuroretinitis, chorioretinitis, uveitis, and retinal vasculitis (20–22), B. henselae has also been reported to cause visceral organ (liver and spleen), cerebral, and cardiac disease (2,23,24).

Golnik et al (5) were among the first to recognize B. henselae (at the time known as Rochalimaea henselae) as the causative agent of neuroretinitis. Intravenous fluorescein angiography has shown that the subretinal fluid and macular exudates are the result of leakage from the surface vasculature of the optic disc.

Increased knowledge of the pathophysiology of B. henselae has provided insight into the clinical manifestations and complications of the infectious process, and a nidus for the improvement in and the development of novel treatment strategies (25,26). It is beyond the scope of this discussion to detail the molecular mechanism by which B. henselae infects humans, but in broad terms, it has a predilection to infect endothelial and erythrocyte cells. Surface specific antigens allow binding to the extracellular matrix and nucleated cells. Once inside the cell, the bacterium avoids the defensive strategy of the host cell, specifically the lysosomal degradation pathway, by residing in vacuoles and introducing anti-endocytic proteins into the cytoplasm (27). In addition, B. henselae infection results in angiogenesis through the production of vascular endothelial growth factor and suppression of pericyte cell proliferation (28).

Back to Top | Article Outline

What Is the Natural History of Bartonella Neuroretinitis?

The natural history of Bartonella neuroretinitis is not known because the published data are composed of case series with relatively small numbers, admixed with treated and untreated patients with neuroretinitis, and various other ocular complications (20–22). Chi et al (29) published the largest series of patients with cat scratch optic neuropathy (62 eyes of 53 patients) and found that the overall visual outcome was excellent at final follow-up with 68% of patients retaining a vision of 20/40 or better. However, many of the patients in the study were treated, thus not allowing for an analysis of the natural history of the disease.

Back to Top | Article Outline

What Is the Evidence to Support Treatment and How Strong Is That Evidence?

A variety of antibiotics have been recommended for B. henselae infection including erythromycin, amoxicillin, gentamicin, rifampin, ceftriaxone, ciprofloxacin, azithromycin, and trimethoprim–sulfamethoxazole (2,25). Although records of case reports and case series have suggested a treatment benefit of antibiotics for patients with CSD, to date, there has only been one randomized controlled trial (RCT) that has evaluated the benefit of antibiotic therapy; and in that study, there was no improved cure rate with treatment (azithromycin) compared with placebo (30).

Published treatment recommendations for Bartonella neuroretinitis should be viewed with caution and if possible self-vetted. For example, in the excellent review article by Rolain et al (2), the authors recommend patients with “retinitis” be treated with doxycycline and rifampin. The quality of evidence and strength of recommendation was given a score of II and A, respectively. Two references were given to support this grade. The first reference is a textbook chapter that states “There is no convincing evidence that antibiotic treatment affects the course of either complicated or uncomplicated CSD in the immunocompetent patient” (31). The second reference is a retrospective case series reporting the visual outcome of 6 patients treated with antibiotics (4- to 6-week course of oral doxycycline [100 mg twice daily] and rifampin [300 mg 2 to 3 times daily]) and one patient who received the same antibiotic regiment and a 2-week course of oral prednisone. The study did not include any untreated patients; therefore, the authors compared the results of their 7 patients with historical data from 2 case reports, a 4-patient case series, and a retrospective study on the therapeutic outcome of 268 patients with CSD some of whom had neuroretinitis, although the exact number of cases were not reported, but “blindness” was noted in 6% of patients (4,32).

It seems clear that the treatment data for Bartonella neuroretinitis is very limited and based solely on data obtained from retrospective, observational open-labeled studies (1). As mentioned earlier, in the largest collection of cat scratch optic neuropathy cases published to date, Chi et al (29) did not find an improved visual outcome in treated (antibiotics, corticosteroids, or both) compared with untreated patients. However, an RCT would be needed to determine whether there is (or is not) a beneficial effect of antibiotic treatment for visual function. If one were to design an RCT using the data from the study of Chi et al, based on the calculations of a two group chi-square test with a 0.050 two-sided significance level with a 90% power to detect the difference between a nontreated group proportion of 0.89 and a treated group proportion of 0.65 (odds ratio of 0.230), the sample size in each group (treatment and placebo) would need to be 63 with total sample size of 126. Given the rarity of Bartonella neuroretinitis, such a recruitment goal for an RCT would be very difficult to accomplish and would require a multicenter study.

Back to Top | Article Outline

What Are the Risks of Treatment?

Both antibiotic and corticosteroid treatments are not innocuous and can be associated with complications, some of which can be severe. It has been estimated that approximately 20% of emergency department visits are because of antibiotic-related adverse events (33). The lax use of antimicrobial agents can result in changes in the normal flora of an individual resulting in disease (e.g., Clostridium difficile colitis) and the development of bacterial resistance (34). The multitude of side effects from corticosteroid treatment is well known to most physicians (e.g., diabetes mellitus, gastric ulcer disease, osteoporosis, immunosuppression, avascular necrosis) resulting in significant morbidity and in some cases death. However, what may not be appreciated is the fact that even young patients receiving short courses of corticosteroid treatment can develop adverse side effects. For example, in the Optic Neuritis Treatment Trial, which enrolled patients between the ages of 18 and 46 years, of the 151 patients treated with intravenous methylprednisolone, 2 patients experienced significant complications of acute pancreatitis and acute psychosis (35).

Back to Top | Article Outline

Rebuttal: Dr. Lee

I agree with Dr. Bhatti that the evidence for treating patients with Bartonella neuroretinitis is sparse, but I don't want to exaggerate the harm. The incidence of community-acquired C. difficile infection in Olmsted County, MN was 9.6 per 100,000 person-years from 1991 to 2005 (36). As for corticosteroids, I would propose limiting treatment to a 2- to 3-week course at most. Although side effects can occur at this dose and duration, a retrospective review of 83 patients treated with at least 1 month of relatively high-dose prednisone and at least 4 months' taper showed a 2-year risk of serious complications of 1% (one patient total, but the authors note that the patient did not follow the prescribed prednisone regimen) (37). It is my belief that a short course of oral antibiotics or oral corticosteroids, in the proper context, has a reasonable safety profile.

Although the article by Chi et al (29) did not find an association between final visual acuity and the use of systemic antibiotic or corticosteroid, there was no randomization. We know that the patients did not have equal visual acuity at baseline. Treated eyes showed worse initial visual acuity than untreated eyes (count fingers vs 20/126 [corticosteroids] and 20/200 vs 20/80 [antibiotics]). We do not know if the groups experienced similar duration of symptoms before initial evaluation. It is unclear which antibiotic was used in each case, and we cannot lump all antibiotics into the same group. I agree with Dr. Bhatti that an RCT is necessary to determine whether a benefit exists for antibiotic treatment or not. Until such a time, we need an outline for treatment. As noted in my initial discussion, although most patients do not require treatment, it is definitely reasonable to consider antibiotics and/or steroids under certain circumstances as listed earlier.

Back to Top | Article Outline

Rebuttal: Dr. Bhatti

To borrow the wonderful aphorism by the great Carl Sagan, “the absence of evidence is not the evidence of absence.” Although I made the argument in my opening remarks that immunocompetent patients with Bartonella neuroretinitis do not need to be treated, I must declare that the lack of clinical data does not mean that one should not treat such a patient. In fact, evidence-based medicine implies the practice of implementing clinical evidence and clinical experience (38).

As expected, Dr. Lee does a marvelous job explaining the reasons for treating a patient with Bartonella neuroretinitis and succinctly reviews the other causes of neuroretinitis and the proper evaluation of a patient with neuroretinitis. However, I respectfully disagree with the statement “there is a compelling rationale for offering treatment for certain patients” for the simple reason that there is a lack of compelling evidence in the literature to offer treatment. It was pointed out that antibiotic treatment of CSD shortens the symptomatic illness (at least some aspects of it) and that in vitro studies have shown B. henselae to be susceptible to a variety of antibiotics. On those 2 points I agree, but I am not convinced that antibiotics have a beneficial effect of visual recovery in patients with Bartonella neuroretinitis. The study of Reed et al (4) is invoked to support the concept that antibiotic therapy helps patients to recover vision quicker and shorten the disease course. As I discussed earlier, this study was a retrospective case series of only 7 patients that was compared with a nebulous historical control group. I do not believe one can use this data to state confidently that the majority of patients with Bartonella neuroretinitis will benefit from antibiotic therapy. Finally, Dr. Lee refers to the study by Bass et al (3), which randomized a total of 29 patients with CSD to either azithromycin (14 patients) or placebo (15 patients) and found a favorable difference in the rate and degree of total lymph node volume decrease in the azithromycin group compared with the placebo group. The more clinically relevant finding of the study was the lack of difference between the 2 groups in any of the measured clinical outcomes, those being fever, malaise/fatigue, anorexia/weight loss, headache, and rash or other skin lesions.

Dr. Lee and I both agree that at a minimum most immunocompetent patients with Bartonella neuroretinitis do not require treatment. Furthermore, I believe prescribing pharmacological (antibiotic or corticosteroid) therapy for any patient with Bartonella neuroretinitis should not be a “knee jerk” response, but rather a decision derived from thoughtful contemplation about one's own (and colleague's) clinical experience, the limited published clinical evidence and the potential harmful side effects of treatment.

Back to Top | Article Outline


Management of Bartonella neuroretinitis is complicated by the rarity of the disorder, known mimickers of optic disc edema with a macular star figure including other infectious and inflammatory diseases, severe papilledema, hypertensive retinopathy, and most importantly an unknown natural history. The decision about when and how to treat should be considered a practice option that is based on careful analysis of the available medical evidence and the characteristics of the individual patient. A reasonable strategy might be to defer treatment in most patients and to consider treatment in patients with profound visual loss, prominent systemic symptoms, and those who are immunocompromised.

Back to Top | Article Outline


1. Purvin V, Sundaram S, Kawasaki A. Neuroretinitis: review of the literature and new observations. J Neuroophthalmol. 2011;31:58–68.
2. Rolain JM, Brougui P, Koehler JE, Maguina C, Dolan MJ, Raoult D. Recommendations for treatment of human infections caused by Bartonella species. Antimicrob Agents Chemother. 2004;48:1921–1933.
3. Bass JW, Freitas BC, Sisler CL, Chan DS, Vincent JM, Person DA, Claybaugh JR, Wittler RR, Weisse RR, Regnery RL, Slater LN. Prospective randomized double blind placebo-controlled evaluation of azithromycin for treatment of cat-scratch disease. Pediatr Infect Dis J. 1998;17:447–452.
4. Reed JB, Scales DK, Wong MT, Lattuada CP, Dolan MJ, Schwab IR. Bartonella henselae neuroretinitis in cat scratch disease: diagnosis, management, and sequelae. Ophthalmology. 1998;105:459–466.
5. Golnik KC, Marotto ME, Fanous MM, Heitter D, King LP, Halpern JI, Holley PH Jr. Ophthalmic manifestations of Rochalimaea species. Am J Ophthalmol. 1994;118:145–151.
6. Conrad DA. Treatment of cat-scratch disease. Curr Opin Pediatr. 2001;13:56–59.
7. George JG, Bradley JC, Kimbrough RC III, Shami MJ. Bartonella quintana associated neuroretinitis. Scand J Infect Dis. 2006;38:127–154.
8. Moreno RJ, Weisman J, Waller S. Neuroretinitis: an unusual presentation of ocular toxoplasmosis. Ann Ophthalmol. 1992;2468–2470.
9. Arruga J, Valentines J, Mauri F, Roca G, Salom R, Rufi G. Neuroretinitis in acquired syphilis. Ophthalmology. 1985;92:262–270.
10. Karma A, Stenborg T, Summanen P, Immonene I, Mikkila H, Seppala I. Long term follow-up of chronic Lyme neuroretinitis. Retina. 1996;16:505–509.
11. Stechschulte SU, Kim RY, Cunningham ET Jr. Tuberculous neuroretinitis. J Neuroophthalmology. 1999;19:201–204.
12. Kosmorsky GS, Prayson R. Primary optic pathway sarcoidosis in a 38-year-old white man. J Neuroophthalmol. 1995;16:188–190.
13. Shoari M, Katz BJ. Recurrent neuroretinitis in an adolescent with ulcerative colitis. J Neuroophthalmol. 2005;25:286–288.
14. Matsuda A, Chin S, Ohashi T. A case of neuroretinitis associated with long-standing polyarteritis nodosa. Ophthalmologica. 1994;208:168–171.
15. Vaphiades MS. Rocky Mountain spotted fever as a cause of macular star figure. J Neuroophthalmol. 2003;23:276–277.
16. Balaskas K, Vaudaux J, Boillat-Blanco N, Guex-Crosier Y. Azithromycin versus sulfadiazine and pyrimethamine for non-vision-threatening toxoplasmic retinochoroiditis: a pilot study. Med Sci Monit. 2012;19:CR296–CR302.
17. Hansmann Y. Treatment and prevention of Lyme disease. Curr Probl Dermatol. 2009;37:111–129.
18. Mosepele M, Mazo D, Cohn J. Bartonella infection in immunocompromised hosts: immunology of vascular infection and vasoproliferation. Clin Dev Immunol. 2012;2012:612809.
19. Breitschwerdt EB, Linder KL, Day MJ, Maggi RG, Chomel BB, Kempf VA. Koch's postulates and the pathogenesis of comparative infectious disease causation associated with Bartonella species. J Comp Pathol. 2013;148:115–125.
20. Kalogeropoulos C, Koumpoulis I, Mentis A, Pappa C, Zafeiropoulos P, Aspiotis M. Bartonella and intraocular inflammation: a series of cases and review of literature. Clin Ophthalmol. 2011;5:817–829.
21. Ormerod LD, Skolnick KA, Menosky MM, Pavan PR, Pon DM. Retinal and choroidal manifestations of cat-scratch disease. Ophthalmology. 1998;105:1024–1031.
22. Solley WA, Martin DF, Newman NJ, King R, Callanan DG, Zacchei T, Wallace RT, Parks DJ, Bridges W, Sternberg P Jr. Cat scratch disease: posterior segment manifestations. Ophthalmology. 1999;106:1546–1553.
23. Podsiadly E, Chmielewski T, Tylewska-Wierzbanowska S. Bartonella henselae and Borrelia burgdorferi infections of the central nervous system. Ann N Y Acad Sci. 2003;990:404–406.
24. Maritsi DN, Zarganis D, Metaxa Z, Papaioannou G, Vartzelis G. Bartonella henselae infection: an uncommon mimicker of autoimmune disease. Case Rep Pediatr. 2013;2013:726826.
25. Breitschwerdt EB, Maggi RG, Chomel BB, Lappin MR. Bartonellosis: an emerging infectious disease of zoonotic importance to animals and human beings. J Vet Emerg Crit Care (San Antonio). 2010;20:8–30.
26. Margileth AM, Hayden GF. Cat scratch disease. From feline affection to human infection. N Engl J Med. 1993;329:53–54.
27. Eicher SC, Dehio C. Bartonella entry mechanisms into mammalian host cells. Cell Microbiol. 2012;14:1166–1173.
28. Varanat M, Maggi RG, Linder KE, Breitschwerdt EB. Infection of human brain vascular pericytes (HBVPs) by Bartonella henselae. Med Microbiol Immunol. 2013;202:143–151.
29. Chi SL, Stinnett S, Eggenberger E, Foroozan R, Golnik K, Lee MS, Bhatti MT. Clinical characteristics in 53 patients with cat scratch optic neuropathy. Ophthalmology. 2012;119:183–187.
30. Prutsky G, Domecq JP, Mori L, Bebko S, Matzumura M, Sabouni A, Shahrour A, Erwin PJ, Boyce TG, Montori VM, Malaga G, Murad MH. Treatment outcomes of human bartonellosis: a systematic review and meta-analysis. Int J Infect Dis. 2013;17:e811–e819.
31. Koehler JE, Relman DA. Bartonella species. In: Yu VL, Merigan TC, Barriere SL, eds. Antimicrobial Therapy and Vaccines. Baltimore, MD: Williams & Wilkins, 1999:578–583.
32. Margileth AM. Antibiotic therapy for cat-scratch disease: clinical study of therapeutic outcome in 268 patients and a review of the literature. Pediatr Infect Dis J. 1992;11:474–478.
33. Shehab N, Patel PR, Srinivasan A, Budnitz DS. Emergency department visits for antibiotic-associated adverse events. Clin Infect Dis. 2008;47:735–743.
34. McDonnell Norms Group. Antibiotic overuse: the influence of social norms. J Am Coll Surg. 2008;207:265–275.
35. Beck RW, Cleary PA, Anderson MM Jr, Keltner JL, Shults WT, Kaufman DI, Buckley EG, Corbett JJ, Kupersmith MJ, Miller NR, et al.. A randomized, controlled trial of corticosteroids in the treatment of acute optic neuritis. The Optic Neuritis Study Group. N Engl J Med. 1992;326:581–588.
36. Khanna S, Pardi DS, Aronson SL, Kammer PP, Orenstein DO, St Sauver JL, Harmsen WS, Zinsmeister AR. The epidemiology of community-acquired Clostridium difficile infection: a population-based study. Am J Gastroenterol. 2012;107:89–95.
37. Bruce BB, Kupersmith MJ. Safety of prednisone for ocular myasthenia gravis. J Neuroophthalmol. 2012;32:212–215.
38. Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS. Evidence based medicine: what it is and what it isn't. BMJ. 1996;312:71–72.
© 2014 by North American Neuro-Ophthalmology Society