INTRODUCTION — Bartonella infections can cause serious morbidity and mortality in people with HIV, particularly those with advanced immunosuppression [1].

This topic will address the diagnosis, treatment, and prevention of Bartonella infections in people with HIV. The epidemiology and clinical manifestations of disease are discussed elsewhere. (See "Epidemiology and clinical manifestations of Bartonella infections in persons with HIV".)

DIAGNOSIS

General background — The diagnosis of Bartonella infections can be challenging in people with HIV and usually requires a combination of tests. The approach to making the diagnosis should consist of obtaining one or more definitive diagnostic tests in combination with serologic testing. Depending on the patient's clinical presentation, tests used to make a definitive diagnosis may include culture of a tissue or blood sample, Bartonella polymerase chain reaction (PCR) testing on a tissue or blood sample, and/or tissue biopsy with histopathologic examination and appropriate staining.

A positive result from more than one type of definitive test provides the strongest evidence for a diagnosis of Bartonella infection. However, since each of the definitive tests has limitations, serologic testing for Bartonella is often used as supportive evidence for infection.

Culture — Bartonella species are fastidious gram-negative bacteria that require specific laboratory conditions to enhance the yield. The likelihood of isolating the organism from blood increases if the blood sample is collected in pediatric or adult isolator tubes (Wampole, Cranbury, New Jersey), or in tubes containing ethylenediaminetetraacetic acid (EDTA). Bartonella species have occasionally been isolated from BACTEC (Becton Dickinson Diagnostic Instrument Systems, Sparks, Maryland) bottles.

Plating blood or tissue samples onto either chocolate agar or heart infusion agar supplemented with 5 percent rabbit blood enhances the yield of isolating organisms in subcultures. Optimally, the microbiology lab should use fresh agar plates and incubate them in 5 percent CO₂ at 35 to 37°C for a minimum of 21 days. Isolation of Bartonella from tissue samples remains very difficult, but some have successfully accomplished this by directly plating tissue homogenates onto agar and cocultivating with a bovine endothelial cell monolayer [2].

Polymerase chain reaction — The use of PCR-based tests have increasingly played an important role in the diagnosis of Bartonella-associated infections, especially since isolating Bartonella species from blood or tissue samples is very difficult [3,4]. For persons with suspected bacteremia, Bartonella PCR testing can be performed on whole blood, plasma, or serum samples. Qualitative Bartonella PCR testing on tissue or blood is available commercially through several laboratories, but not all of the laboratories offer a PCR test that differentiates Bartonella quintana from Bartonella henselae [5].

Histopathology — Standard hematoxylin and eosin staining of bacillary angiomatosis (BA) lesions from any site characteristically shows lobular vascular proliferations composed of rounded vessels lined by variably protuberant plump endothelial cells (picture 1) [6]. In addition, clusters of neutrophils, neutrophilic debris, and lymphocytes are scattered throughout the lesions, especially around eosinophilic granular aggregates. If possible, Warthin-Starry silver staining should be performed, which may reveal masses of small, dark-staining bacteria (picture 2). Electron microscopic examination, if performed, shows pleomorphic bacilli with a trilaminar wall (picture 3) [7].

Histopathologic findings in peliosis hepatis and splenitis typically consist of cystic blood-filled spaces and fibromyxoid stroma that contains a mixture of inflammatory cells, dilated capillaries, and granular purple material. Warthin-Starry staining of these lesions demonstrates findings similar to BA. (See "Peliosis hepatis".)

Serology — Serologic testing is often used as supportive, but not definitive, evidence for infection.

Two serologic methods, indirect fluorescence assay (IFA) and enzyme immunosorbent assay (EIA), have been tested for the diagnosis of B. henselae and B. quintana infections. Most of the data with these tests have come from studies involving persons with cat scratch disease [8,9]. It is not clear whether persons with HIV develop antibodies to the Bartonella organisms in the same fashion as immunocompetent individuals. In one serologic study in Baltimore that involved persons who inject drugs, HIV infection and low CD4 counts were associated with a decreased probability of seropositivity to Bartonella [10]. Another study demonstrated that nearly one-fourth of persons with late-stage HIV not on antiretroviral therapy who have Bartonella infection documented by culture never developed serum antibodies [11].

Problems with the B. henselae IFA serology tests include:

●Significant cross reactivity occurs at the species level between B. henselae and B. quintana, especially for IgG assays.

●The sensitivity of the test does not appear to be optimal, especially with IgG assays.

●The prevalence of positive Bartonella serology in the general population is 4 to 6 percent, which creates difficulties with false positive tests.

In the United States, serologic tests for Bartonella are available through multiple commercial laboratories. Although both the IFA and EIA serologic assays have been studied, commercially available assays typically use the IFA assay, and offer testing for both IgM and IgG. Most laboratories that perform Bartonella IFA tests report separate results for B. quintana and B. henselae, but the significant cross-reactivity limits the value of species-specific results.

In general, IFA IgG titers <1:64 suggest the person undergoing testing does not have a current Bartonella infection, but up to 25 percent of persons with HIV who have advanced immunosuppression may never develop a positive serologic test [11]. Titers <1:64 could represent past infection. Titers >1:64 but <1:256 represent possible Bartonella infection, and repeat testing in 10 to 14 days is generally recommended. Titers >1:256 strongly suggest active or recent infection.

Since data about Bartonella serologic testing in people with HIV are sparse, these tests should not stand alone as a means to diagnose Bartonella infections. A positive serology can be used as supportive information in the appropriate clinical setting. However, a convincing diagnosis of Bartonella infection in persons with HIV generally requires a positive culture or PCR test, or a biopsy showing histopathologic findings consistent with Bartonella infection.

THERAPY

Antimicrobial regimen — All persons with HIV who are diagnosed with Bartonella infection should receive antibiotic therapy [1,12]. No randomized controlled trials have evaluated the optimal treatment regimen for Bartonella in persons with HIV [1]. However, in a summary of data from approximately 50 patients at San Francisco General Hospital, persons with HIV generally responded well to prolonged courses of either erythromycin or doxycycline [13].

Our approach to treatment is based upon panel guidelines [1,12] and clinical experience. The optimal duration of treatment for people with HIV has not been established, and the duration should be extended in those who have not had adequate clinical improvement.

●Patients with bacillary angiomatosis (BA), peliosis hepatitis, or osteomyelitis should receive doxycycline (100 mg orally [PO] or intravenously [IV] twice daily) or erythromycin (500 mg PO or IV four times daily). We prefer doxycycline rather than erythromycin due to more favorable dosing schedule and fewer drug interactions.

•Patients with BA should receive oral therapy for three months.

•For those with peliosis hepatitis and/or osteomyelitis, we prefer to initiate intravenous (IV) therapy; such patients can transition to oral therapy after clinical improvement. The total duration of therapy should be four months.

●Patients with central nervous system disease should receive doxycycline (100 mg PO or IV twice daily). The duration of therapy should be three to four months. Some experts also add rifampin (300 mg PO or IV twice daily) for the treatment of such patients [1], although there are no data to support rifampin use to treat central nervous system disease in persons with HIV. In addition, rifampin can cause significant drug interactions, including with antiretroviral medications. (See "Overview of antiretroviral agents used to treat HIV".)

●For patients with bacteremia, we treat with doxycycline (100 mg PO or IV twice daily) plus gentamicin (1 mg/kg IV every eight hours) [14]. The gentamicin dose should be adjusted to achieve peak serum concentrations of 3 to 4 mcg/mL and trough serum concentrations of <1 mcg/mL (see "Dosing and administration of parenteral aminoglycosides"). Gentamicin is used only for the first 14 days; the duration of therapy for doxycycline is three months.

●For patients with life-threatening infection (eg, hypotension or requiring care in the intensive care unit), we add rifampin (300 mg IV every 12 hours) to doxycycline (100 mg IV every 12 hours). For patients with life-threatening infection who are unable to receive doxycycline, erythromycin 500 mg IV every six hours or azithromycin 500 mg IV once daily can be used as an alternative. The patient can transition from intravenous to oral therapy after becoming clinically stable.

●The treatment of Bartonella bacteremia and endocarditis in persons with HIV is the same as for HIV-seronegative patients, except that we prefer a longer duration of therapy (three months) for those with HIV. A detailed discussion of Bartonella endocarditis is found elsewhere. (See "Endocarditis caused by Bartonella".)

●Azithromycin (500 mg once daily) or clarithromycin (500 mg twice daily) can be used as an alternative therapy for patients who cannot take doxycycline or erythromycin. However, the macrolides are likely inferior to doxycycline for the treatment of central nervous system disease.

●Trimethoprim-sulfamethoxazole, ciprofloxacin, penicillins, and first generation cephalosporins do not have reliable activity, and thus are not recommended.

Monitoring response to therapy — A baseline Bartonella IgG should be checked prior to starting antimicrobial therapy as part of the initial diagnostic evaluation. Expert guidelines for people with HIV recommend checking sequential Bartonella IgG titers every six to eight weeks to monitor the initial response to therapy, with the goal of achieving at least a four-fold decline in Bartonella IgG titer [1]. A more detailed discussion of duration of therapy is found above. (See 'Therapy' above.)

There are no studies that have addressed the optimal approach to patients who fail to respond to therapy. In such patients, we add rifampin if it was not used in the original treatment regimen; we also consider adding gentamicin for 14 days for patients with peliosis hepatis.

Suppressive therapy — We agree with expert guidelines that recommend administering suppressive therapy with doxycycline or a macrolide to patients who have a relapse of disease after receiving at least three months of therapy [1]. For patients receiving suppressive therapy, we recommend using the same doses of doxycycline or a macrolide as the ones used for treatment. The patient can discontinue suppressive therapy if they have received at least three months of therapy and their CD4 count is >200 cells/microL for at least six months.

IMMUNE RECONSTITUTION INFLAMMATORY SYNDROME — On rare occasion, persons with HIV and Bartonella infection have developed a significant worsening of Bartonella infection after initiating antiretroviral therapy [15,16]. This immune reconstitution inflammatory syndrome (IRIS) occurs as a consequence of rapid up-regulation in the immune system and, if severe, may require corticosteroid therapy. (See "Immune reconstitution inflammatory syndrome".)

WHEN TO INITIATE ANTIRETROVIRAL THERAPY — In general, antiretroviral therapy (ART) should be administered to people with HIV who are not receiving ART when Bartonella infection is diagnosed. When selecting a regimen, it is important to assess for drug interactions, particularly if rifampin is used. (See "Selecting antiretroviral regimens for treatment-naïve persons with HIV-1: General approach".)

However, the exact timing for when to initiate ART depends on the clinical presentation of Bartonella infection. We agree with guidelines that suggest deferring ART two to four weeks after starting antimicrobial treatment for Bartonella in patients who have central nervous system or ophthalmic disease [1]. The rationale for this approach is based upon concern that developing an immune reconstitution inflammatory syndrome (IRIS) that may lead to serious complications related to the Bartonella infection. Although there are only rare reports describing IRIS in patients with Bartonella infection [15,16], poor clinical outcomes related to IRIS have been demonstrated in patients with other opportunistic infections that affect the central nervous system and the eye, such as cytomegalovirus and toxoplasmosis. (See "Immune reconstitution inflammatory syndrome" and "Treatment of AIDS-related cytomegalovirus retinitis", section on 'CMV immune reconstitution inflammatory syndromes' and "Toxoplasmosis in patients with HIV", section on 'Toxoplasmosis and immune reconstitution syndrome (IRIS)'.)

We do not delay the initiation of ART in patients presenting with other manifestations of Bartonella (eg, skin lesions, endocarditis) since the risks associated with an IRIS-like syndrome would be less severe in these conditions. (See "Immune reconstitution inflammatory syndrome".)

PREVENTION — We do not administer antimicrobial agents specifically to prevent Bartonella infection in people with HIV [1]. A single retrospective, case-control study reported a protective effect against Bartonella infection in patients receiving a macrolide or a rifamycin for prophylaxis against disseminated Mycobacterium avium complex [17]. Despite this finding, we do not believe prophylactic therapy specific for Bartonella infection is indicated given the low incidence of Bartonella infection in persons with HIV and the lack of prospective data.

However, we counsel patients, especially those with a CD4 count <200 cells/microL, on how to decrease the risk of developing Bartonella. Since the development of B. henselae infection is related to cat exposure, this includes:

●Avoiding cat scratches

●Controlling flea infestations

●Avoiding acquisition of an unhealthy cat or a cat that is younger than one year old

Homeless or intermittently housed persons with HIV should receive:

●Education regarding prevention of lice infestation

●Treatment of lice infestation to minimize their risk of developing systemic B. quintana infection

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Opportunistic infections in HIV-infected adults and adolescents".)

SUMMARY AND RECOMMENDATIONS

●Bartonella infections can cause serious morbidity and mortality in persons with HIV, particularly those with advanced immunosuppression. However, the diagnosis of Bartonella infections can be challenging in persons with HIV and usually requires a combination of tests. (See 'General background' above.)

●Although isolation of this organism makes a definitive diagnosis, this is infrequently accomplished due to the fastidious nature of this organism. (See 'Culture' above.)

●Polymerase chain reaction (PCR) testing plays an important role in the diagnosis of Bartonella because of the difficulty in isolating this organism from tissue samples. The Bartonella PCR test is available from many laboratories, but not all offer a PCR assay that will distinguish among the different Bartonella species. (See 'Polymerase chain reaction' above.)

●The diagnosis of bartonellosis can be made through histopathologic examination of biopsy tissue. In patients with bacillary angiomatosis, lesions are characterized by vascular proliferation, and Warthin-Starry staining usually demonstrates numerous bacilli. Histopathologic findings in peliosis hepatis and splenitis typically consist of cystic blood-filled spaces and fibromyxoid stroma that contain a mixture of inflammatory cells and dilated capillaries. (See 'Histopathology' above.)

●Serologic testing is often used as supportive, but not definitive, evidence for infection. (See 'Serology' above.)

●All persons with HIV who are diagnosed with Bartonella infection should receive antibiotic therapy. Although no randomized controlled trial data are available, observational studies suggest that patients generally respond well to prolonged courses of either doxycycline or erythromycin. (See 'Therapy' above.)

●For patients with bacteremia, we administer doxycycline (100 mg orally or intravenously [IV] twice daily) plus gentamicin (1 mg/kg IV every eight hours) [14]. The gentamicin is used only for the first 14 days; doxycycline should be given for three months. (See 'Therapy' above.)

●Long-term suppressive therapy with a macrolide or doxycycline is recommended for patients with a history of relapse of Bartonella infection, and treatment should be continued until the CD4 count has increased to >200 cells/microL for at least six months. (See 'Suppressive therapy' above.)

●In patients who are not receiving antiretroviral therapy (ART), and who have central nervous system or ophthalmic disease, ART should be initiated two to four weeks after starting antimicrobial treatment for Bartonella to minimize the risk of a severe immune reconstitution syndrome reaction. We do not delay initiation of ART in patients presenting with other manifestations of Bartonella. (See 'When to initiate antiretroviral therapy' above.)

●In people with HIV, we do not administer antimicrobial agents specifically to prevent Bartonella infection. However, we provide counseling on how to decrease the risk of exposure. (See 'Prevention' above.)

ACKNOWLEDGMENT — We are saddened by the death of John G Bartlett, MD, who passed away in January 2021. UpToDate gratefully acknowledges Dr. Bartlett's role as section editor on this topic, his tenure as the founding Editor-in-Chief for UpToDate in Infectious Diseases, and his dedicated and longstanding involvement with the UpToDate program.