Guidelines for the Prevention and Treatment of Opportunistic Infections in Adults and Adolescents With HIV

Epidemiology

Bartonella species cause infections that include cat scratch disease, retinitis, trench fever, relapsing bacteremia, culture-negative endocarditis, bacillary angiomatosis (BA), and bacillary peliosis hepatis.¹ The latter two manifestations occur almost exclusively in individuals who are immunocompromised. Thirty-seven species and three subspecies of Bartonella have been described and are officially recognized (see Bartonella on the List of Prokaryotic Names with Standing in Nomenclature); fourteen of these Bartonella species have been implicated in human infections.

BA most often occurs late in HIV infection² in patients with median CD4 T lymphocyte (CD4) cell counts <50 cells/mm³. In people with HIV, bartonellosis is often a chronic illness, lasting for months to more than a year, with BA lesions and intermittent bacteremia. Development of BA lesions caused by B. henselae is statistically linked to cat exposure in people with HIV.² In contrast, BA caused by B. quintana is associated with body louse infestation and homelessness.² The body louse serves as the vector of B. quintana to humans. To avoid exposure to B. quintana, people with HIV should avoid body lice exposure and have prompt eradication of lice if infestation occurs. The cat flea is the vector of B. henselae in cats. Cats are the most common vector (via a scratch) responsible for transmitting B. henselae to humans, most likely when their claws become contaminated with feces from B. henselae-infected fleas. In some areas of the United States, the prevalence of B. henselae bacteremia in pet cats approaches 50%;³ infection is more common among kittens and feral cat populations. Controlling cat flea infestation and avoiding cat scratches are therefore critical strategies for preventing B. henselae infections in people with HIV.

Clinical Manifestations

BA lesions have been associated with nearly every organ system, but cutaneous lesions are the most readily identified. These lesions can be clinically indistinguishable from Kaposi sarcoma, pyogenic granuloma, and other skin conditions. BA also can cause subcutaneous nodules. Osteomyelitis is usually caused by B. quintana, and only B. henselae causes bacillary peliosis hepatis.² Although isolated organs can appear to be the principal focus of disease, BA represents a hematogenously disseminated infection, and systemic symptoms of fever, night sweats, and weight loss often accompany BA. Bartonella infection is a major cause of unexplained fever in patients with advanced HIV and should be considered in the differential diagnosis of patients with CD4 counts <100 cells/mm³ and fever.⁴ Bartonella is a frequent cause of culture-negative endocarditis in immunocompetent and immunocompromised humans and is most commonly caused by B. quintana, less frequently by B. henselae, and rarely by other Bartonella species.⁵ Immune complex disease (such as glomerulonephritis) may complicate endocarditis or other systemic Bartonella infections; assessment for immune complex formation may be warranted in such cases so that nephrotoxic agents can be avoided.

Diagnosis

Diagnosis of BA can be confirmed by histopathologic examination of biopsied tissue.⁶ BA lesions are characterized by vascular proliferation, and a modified silver stain (such as Warthin-Starry stain) usually demonstrates numerous bacilli. Tissue Gram staining and acid-fast staining are negative.

A well-characterized indirect fluorescent antibody (IFA) serologic test was developed at the Centers for Disease Control and Prevention (CDC)⁷ and is available at the CDC Infectious Diseases Laboratories. In addition, several private laboratories offer IFA serological testing, but the performance characteristics of these tests have not been validated for people with HIV. In immunocompetent patients, anti-Bartonella antibodies might not be detectable for 6 weeks after acute infection; in contrast, by the time Bartonella infection is suspected in patients with late-stage HIV infection, they usually have been infected with Bartonella for months or even >1 year. However, as many as 25% of Bartonella culture-positive patients never develop antibodies in the setting of advanced HIV infection.⁴ In those patients who do develop anti-Bartonella antibodies, monitoring of antibody levels can be useful in following treatment response of Bartonella infection to antibiotics, reflecting resolution⁸ or recrudescence. Because of interlaboratory variability, longitudinal testing should be conducted at the same laboratory to enable direct comparison of titers over time.

Because of their fastidious nature, Bartonella organisms can be isolated only with difficulty from blood (drawn into ethylenediaminetetraacetic acid [EDTA] tubes, centrifuged, and then plated directly onto fresh chocolate agar). Bartonella has been cultured directly from tissue in only a few laboratories.² Removing samples from blood culture bottles after 8 days of incubation, followed by staining with acridine orange, has facilitated identification and subsequent culture of Bartonella species.⁹ Additionally, the CDC can perform polymerase chain reaction (PCR) amplification with universal and/or specific primers to detect Bartonella in EDTA blood samples (see Bartonella quintana Molecular Detection); these molecular detection tests also are increasingly available through private laboratories. Finally, molecular detection of Bartonella in BA skin lesions or other vascular lesions, lymph nodes, or resected cardiac valves from unfixed tissue biopsy samples (at the University of Washington) or from formalin-fixed tissue (at the CDC Infectious Disease Pathology Branch) can be performed.^8,10 Bartonella species may also be detected from blood or plasma using metagenomic next generation sequencing.^11-13 Clinicians should be aware that results from the CDC may take longer—several weeks to months—for serologic and molecular testing, respectively, compared with some private laboratories. A notable update was published in the 2023 Duke-ISCVID Criteria for Infective Endocarditis, indicating that an IFA immunoglobulin G (IgG) titer of ≥1:800 for B. quintana or B. henselae or identification of a Bartonella sp. by PCR or other nucleic acid–based techniques (including metagenomic sequencing) from blood are now considered major criteria for the diagnosis of Bartonella endocarditis.¹⁴

In summary, diagnosis of bartonellosis may require multiple testing modalities, including serologic testing (which is the most accessible test, and when positive, is helpful both for diagnosis and subsequent monitoring of treatment response), histopathology, and, especially, molecular testing for biopsied or resected tissue (e.g., BA lesion tissue or heart valve tissue).

Preventing Exposure

People with HIV, specifically those who are severely immunocompromised (CD4 counts <100 cells/mm³), are at high risk of severe disease when infected by B. quintana or B. henselae. The major risk factors for acquisition of B. henselae are contact with cats infested with fleas and receiving cat scratches. Immunocompromised individuals should consider the potential risks of cat ownership (AIII). People with HIV who want cats should acquire animals that are older than 1 year of age and in good health (BII). Cats should be acquired from a known environment, have a documented health history, and be free of fleas. Stray cats and cats with flea infestation should be avoided. Declawing is not advised, but individuals with HIV should avoid rough play with cats and situations in which scratches are likely (AII). People with HIV should avoid contact with flea feces (i.e., flea dirt), and any cat-associated wound should be washed promptly with soap and water (BIII). Care of cats should include a comprehensive, ongoing flea-control program under the supervision of a veterinarian (BIII). No evidence indicates any benefits to cats or their owners from routine culture or serologic testing of the pet for Bartonella infection or from antibiotic treatment of healthy, serologically positive cats (BII). The major risk factor for B. quintana infection is body lice infestation. People with HIV who are experiencing homelessness or are in marginal housing should be informed that body louse infestation can be associated with serious illness and should be provided with appropriate measures to eradicate body lice, if present (AII). Regardless of CD4 count, people with both HIV and solid organ transplantation may be at risk of developing more severe Bartonella infections, similar to transplant recipients without HIV.¹⁵

Preventing Disease

Primary chemoprophylaxis for Bartonella-associated disease is not recommended (BIII). However, note that in a retrospective case-control study, use of a macrolide (such as for Mycobacterium avium complex prophylaxis) was protective against developing Bartonella infection.²

Treating Disease

All patients with HIV and Bartonella infection should receive antibiotic treatment (AII). No randomized, controlled clinical trials have evaluated antimicrobial treatment of bartonellosis in patients with HIV. Erythromycin and doxycycline have been used successfully to treat BA, peliosis hepatis, bacteremia, and osteomyelitis; either drug is considered first-line treatment for bartonellosis on the basis of reported experience in case series (AII).^1,2 Anecdotal and limited published case reports¹⁶ suggest that other macrolide antibiotics (such as azithromycin or clarithromycin) are effective in treating Bartonella infections in patients with HIV and may be better tolerated than erythromycin; either of these can be an alternative therapy for Bartonella infections (except for endocarditis or central nervous system [CNS] infections) (BIII). Therapy should be administered for at least 3 months (AII). Doxycycline, preferably in combination with a rifamycin class antibiotic, is the treatment of choice for bartonellosis infection involving the CNS (AIII). For severe Bartonella infections (i.e., patients with multifocal disease or evidence of clinical decompensation), combination therapy using erythromycin or doxycycline with a rifamycin class antibiotic is recommended (BIII); intravenous therapy may be needed initially (AIII). Treatment of Bartonella endocarditis should include doxycycline with the addition of a rifamycin class antibiotic for a minimum of 6 weeks (BII). Doxycycline for 6 weeks plus gentamicin for the first 2 weeks may also be considered but is less preferred due to the intrinsic nephrotoxicity of gentamicin and the frequency of vasculitis-induced renal dysfunction complicating Bartonella endocarditis (BII).¹⁷

Penicillins and first-generation cephalosporins have no in vivo activity and should not be used for treatment of bartonellosis (AII).¹⁸ Bartonella species have been isolated from patients with HIV during documented treatment or prophylaxis with trimethoprim-sulfamethoxazole (TMP-SMX);² quinolones and TMP-SMX also have variable in vitro activity and an inconsistent clinical response in case reports and are not recommended (AIII).

Monitoring of Response to Therapy and Adverse Effects (Including IRIS)

The potential exists for immune reconstitution inflammatory syndrome (IRIS) in association with bartonellosis treatment and initiation of antiretroviral therapy (ART) in people with HIV. In ART-naive patients, ART generally can be initiated at the same time as Bartonella-directed treatment; however, patients with Bartonella CNS or ophthalmic lesions probably should be treated with doxycycline and a rifamycin class antibiotic for 2 to 4 weeks before instituting ART (CIII).

Because of the propensity for relapse of Bartonella infection, patients should have anti-Bartonella IFA IgG antibody titers checked at the time of diagnosis (Note: It is important to specify to the receiving lab that the sample must be diluted to endpoint.) and, if positive, should be followed with sequential endpoint titers every 6 to 8 weeks during treatment, preferably until at least a fourfold decrease is documented (CIII).⁸ Patients treated with oral doxycycline should be cautioned about pill-associated esophagitis and photosensitivity. Adverse effects associated with macrolides include nausea, vomiting, abdominal pain, and elevations of liver transaminase levels; potential QT interval prolongation also should be considered. Serious side effects can occur during treatment with rifamycin class antibiotics, including hypersensitivity reactions (thrombocytopenia, interstitial nephritis, and hemolytic anemia) and hepatitis. Administration of rifamycin class antibiotics strongly induces the cytochrome P450 enzyme system, which is an important consideration when other medications, including many antiretroviral drugs, are taken simultaneously.

Managing Treatment Failure

Relapse of Bartonella infections occurs frequently, especially in patients with BA. Among patients who fail to respond to initial treatment, switching to a different preferred regimen (for example, from doxycycline to erythromycin) may be considered, again with treatment duration of ≥3 months (AIII). For severe infections, the addition of a rifamycin class antibiotic is indicated (AIII). For patients with positive or increasing antibody titers, but with clinical improvement, treatment should continue until at least a fourfold decrease in the antibody titers is documented (CIII).⁸

Preventing Recurrence

After a primary course of treatment (minimum of 3 months), treatment may be discontinued, with close monitoring for evidence of relapse (e.g., symptoms, increase in antibody titers).

If a relapse occurs, an additional course of treatment is recommended, followed by long-term suppression of infection with doxycycline or a macrolide (AIII).

Long-term suppression can be discontinued after the patient has received at least 3 to 4 months of therapy and when the CD4 count remains >200 cells/mm³ on effective ART for ≥6 months (CIII).⁸ Some specialists would discontinue therapy only if the Bartonella titers also have decreased at least fourfold (CIII).

Special Considerations During Pregnancy

Infection with B. bacilliformis in immunocompetent patients during pregnancy has been associated with increased complications and risk of death, but no data are available on the effect of B. quintana or B. henselae infection during pregnancy.

The approach to diagnosis of Bartonella infections in pregnant people is the same as in nonpregnant people. Erythromycin treatment (or an alternative macrolide) should be used as first-line therapy (AIII) rather than tetracyclines (such as doxycycline) during pregnancy because of the increased risk of hepatotoxicity and the accumulation of tetracycline in fetal teeth and bones, resulting in dark, permanent staining of fetal teeth. Third-generation cephalosporins, such as ceftizoxime¹⁹ or ceftriaxone, may have efficacy against Bartonella in pregnant people with HIV, but it should be considered second-line therapy after a macrolide. First- and second-generation cephalosporins are not recommended because of their lack of efficacy against Bartonella (AII).

References

1. Rose SR, Koehler J. Bartonella, including cat-scratch disease. In: Mandell, Douglas, and Bennett's Principles and Practice of Infectious Disease. J. E. Bennett, R. Dolin and M. J. Blaser. 9th ed. Philadelphia, PA: Elsevier Press; 2020.
2. Koehler JE, Sanchez MA, Garrido CS, et al. Molecular epidemiology of bartonella infections in patients with bacillary angiomatosis-peliosis. N Engl J Med. 1997;337(26):1876-1883. Available at: https://www.pubmed.ncbi.nlm.nih.gov/9407154.
3. Koehler JE, Glaser CA, Tappero JW. Rochalimaea henselae infection. A new zoonosis with the domestic cat as reservoir. JAMA. 1994;271(7):531-535. Available at: https://www.pubmed.ncbi.nlm.nih.gov/8301768.
4. Koehler JE, Sanchez MA, Tye S, et al. Prevalence of Bartonella infection among human immunodeficiency virus-infected patients with fever. Clin Infect Dis. 2003;37(4):559-566. Available at: https://www.pubmed.ncbi.nlm.nih.gov/12905141.
5. Fournier PE, Thuny F, Richet H, et al. Comprehensive diagnostic strategy for blood culture-negative endocarditis: a prospective study of 819 new cases. Clin Infect Dis. 2010;51(2):131-140. Available at: https://www.pubmed.ncbi.nlm.nih.gov/20540619.
6. LeBoit PE, Berger TG, Egbert BM, Beckstead JH, Yen TS, Stoler MH. Bacillary angiomatosis. The histopathology and differential diagnosis of a pseudoneoplastic infection in patients with human immunodeficiency virus disease. Am J Surg Pathol. 1989;13(11):909-920. Available at: https://www.pubmed.ncbi.nlm.nih.gov/2802010.
7. Regnery RL, Olson JG, Perkins BA, Bibb W. Serological response to "Rochalimaea henselae" antigen in suspected cat-scratch disease. Lancet. 1992;339(8807):1443-1445. Available at: https://www.pubmed.ncbi.nlm.nih.gov/1351130.
8. Lee SA, Plett SK, Luetkemeyer AF, et al. Bartonella quintana aortitis in a man with AIDS, diagnosed by needle biopsy and 16S rRNA gene amplification. J Clin Microbiol. 2015;53(8):2773-2776. Available at: https://www.pubmed.ncbi.nlm.nih.gov/26063867.
9. Larson AM, Dougherty MJ, Nowowiejski DJ, et al. Detection of Bartonella (Rochalimaea) quintana by routine acridine orange staining of broth blood cultures. J Clin Microbiol. 1994;32(6):1492-1496. Available at: https://www.pubmed.ncbi.nlm.nih.gov/7521357.
10. Dumler JS, Carroll KS, Patel R. Bartonella. In: Manual of Clinical Microbiology. K. C. Carroll, M. A. Pfaller and M. Landry. 12th ed. Washington, DC: American Society for Microbiology; 2019.
11. Li M, Yan K, Jia P, Wei E, Wang H. Metagenomic next-generation sequencing may assist diagnosis of cat-scratch disease. Front Cell Infect Microbiol. 2022;12:7. Available at: https://www.pubmed.ncbi.nlm.nih.gov/36189365.
12. Flurin L, Wolf MJ, Fisher CR, et al. Pathogen detection in infective endocarditis using targeted metagenomics on whole blood and plasma: a prospective pilot study. J Clin Microbiol. 2022;60(9):e0062122. Available at: https://www.pubmed.ncbi.nlm.nih.gov/36040200.
13. Downey RD, Russo SM, Hauger SB, et al. Identification of an emergent pathogen, Bartonella vinsonii, using next-generation sequencing in a patient with culture-negative endocarditis. J Pediatric Infect Dis Soc. 2021;10(2):213-216. Available at: https://www.pubmed.ncbi.nlm.nih.gov/32092135.
14. Fowler VG, Durack DT, Selton-Suty C, et al. The 2023 Duke-ISCVID Criteria for infective endocarditis: updating the modified Duke Criteria. Clin Infect Dis. 2023. Available at: https://www.pubmed.ncbi.nlm.nih.gov/37138445.
15. Psarros G, Riddell Jt, Gandhi T, Kauffman CA, Cinti SK. Bartonella henselae infections in solid organ transplant recipients: report of 5 cases and review of the literature. Medicine (Baltimore). 2012;91(2):111-121. Available at: https://www.pubmed.ncbi.nlm.nih.gov/22391473.
16. Guerra LG, Neira CJ, Boman D, et al. Rapid response of AIDS-related bacillary angiomatosis to azithromycin. Clin Infect Dis. 1993;17(2):264-266. Available at: https://www.pubmed.ncbi.nlm.nih.gov/8399879.
17. Raybould JE, Raybould AL, Morales MK, et al. Bartonella endocarditis and Pauci-immune glomerulonephritis: a case report and review of the literature. Infect Dis Clin Pract (Baltim Md). 2016;24(5):254-260. Available at: https://www.pubmed.ncbi.nlm.nih.gov/27885316.
18. Koehler JE, LeBoit PE, Egbert BM, Berger TG. Cutaneous vascular lesions and disseminated cat-scratch disease in patients with the acquired immunodeficiency syndrome (AIDS) and AIDS-related complex. Ann Intern Med. 1988;109(6):449-455. Available at: https://www.pubmed.ncbi.nlm.nih.gov/3415105.
19. Riley LE, Tuomala RE. Bacillary angiomatosis in a pregnant patient with acquired immunodeficiency syndrome. Obstet Gynecol. 1992;79(5, Part 2):818-819. Available at: https://www.pubmed.ncbi.nlm.nih.gov/1565376.