Co-infection Bartonella

Bartonellosis is an emerging infectious disease that can develop into a chronic and sometimes serious infection. This infection can manifest itself systemically and involves multiple organ systems including the nervous system. Its less benign presentation is called cat scratch fever. It is thought that cat scratch fever is a self limited infection but new research is showing that in a subset of patients the bacteria can chronically persist over years or decades.

Bartonella are spread through various arthropod insects such as fleas, lice, biting flies, mosquitoes and mites. The status of ticks as a competent vector remains questionable. People run the risk of contracting the disease through contact with pets such as dogs and cats or possibly insect bites.

The disease is understood poorly and symptoms caused by a Bartonella infection might get misdiagnosed. Because few doctors understand this infection, or reduce the presentation of the infection to what we refer to as cat scratch fever, it is easy to think it is a new or a rare disease. Neither is true. The disease has been described for centuries, but the causative agent, the bacterial organism Bartonella, has only been identified recently.

Increasing research

Only in the last 20 years have scientists been looking closely at the organism in the laboratory. Analysis showed that Bartonella quintana was a common co-infection in people who died during the Plague epidemic in the 11th century and those who died of Typhoid fever during the early 19th century. Bartonella is an opportunistic co-infection.

Carrión’s disease

Carrión’s disease has been named after Daniel Alcides Carrión Garcia who was a medical student. Carrión described the disease and proved during a lethal experiment that the chronic form of the disease (verruga peruana) and the acute phase (Oroya fever) were one and the same disease.

He innoculated himself with the bacteria from a wart of a 14 year old patient, proving that both clinical pictures were linked to the same cause. The causative agent and way of transmission of the disease were unknown at that time. Eventually Carrión himself would die of the disease with which he infected himself as an experiment. The infectious agent of this infectious disease, Bartonella bacilliformis, was discovered in the early 20th century.

Cat scratch fever

A disease similar to cat scratch fever was first described in 1889 by Henry Perinaud. The clinical syndrome was described and named ‘cat scratch fever’ by Robert Debré in 1950. In 1990, serology proved that the disease was caused by Bartonella henselae. The infection was thought to be self-limited and benign but an increasing amount of case reports show that the infection can become chronic in a subset of patients. Cat scratch fever is a term that is insufficient to describe this condition. Bartonellosis would be a better term.

Trench fever

This disease was described in detail in soldiers who fought in the first World War. The causative agent, Bartonella quintana, was first found in 1961. The research on Bartonella species generally increased around 1990, making it possible to classify the Quintana bacteria to family and genus.

Other Bartonella species

In the last 15 years, 30 different Bartonella species have been discovered. 8-12 of these are able to cause disease in humans. New Bartonella species are found annually. Some researchers therefore argue that the Bartonella genus includes a growing group of bacteria.

Many of these bacteria cause diseases such as Carrion’s disease, cat scratch fever and trench fever. These old terms are slowly abandoned for a more accurate description: ‘Bartonellosis.’

Outdated information

Bartonella is still insufficiently understood and physicians work with outdated information. The focus of transmission is still on cats and lice. As a result, Bartonella is not thought of in people who could have been infected in another way. The disease is often thought to be a benign and self-limited condition known as cat scratch fever. Many patients report that they developed chronic disease as a result of this infection which is reflected in the case reports collected in the references at the end of this article.

One of the tactics of Bartonella is to infect the salivary glands to promote transmission. Through this mechanism, it can be transmitted by a variety of pets. 10% of healthy dogs and 27% of diseased dogs may be infected with Bartonella henselae. In dog saliva 4 other Bartonella species are found that are all pathogenic in humans.

Bartonella henselae is also found in rats, mice, whales, porpoises and dolphins. Research shows that Bartonella organisms are more widely spread among mammals than was previously thought. Also, Bartonella organisms are easily transmitted through a variety of arthropod vectors such as fleas, lice, biting flies, mosquitoes and mites. Some researchers believe ticks might be a competent vector too but the status of Bartonella as a tick borne infection remains questionable.

Bartonella genus

Bartonella are slowly dividing, non motile rod-shaped bacteria. They are picky in nutrition (fastidious), like oxygen, need 22 hours to reproduce, they do not form spores and they live inside cells. (Although they can also aggregate in the extracellular environment) They are very small: 2 to 3 microns long and 0.2 to 0.5 microns wide.


Bartonella belongs to the group of proteobacteria. The proteobacteria are again divided into six groups: alpha, beta, gamma, delta, epsilon and zeta proteobacteria. Interestingly, Ehrlichia, Anaplasma and Rickettsia belong to this group and can all be transmitted by ticks.

The group of proteobacteria contains many species that are pathogenic to people and are becoming increasingly resistant to antibiotics: Klebsiella spp., E. Coli, Vibrio Cholerae, Pseudomonas spp. , Salmonella spp., Shigella spp. and Yersinia spp.

Gram staining

Bartonella is a gram negative organism. Gram-negative bacteria are more difficult to treat than gram-positive bacteria. Important is the difference in cellular structure. The external membrane of a bacterium is called the cell wall. The internal part is called cytoplasm. A gram negative bacterium has a second cell wall called the outer membrane. The space between these two cell walls is called periplasm. Gram-positive bacteria do not have double cell walls. Although the single cell wall is thicker, these bacteria are generally easier to treat.

Probably most important is its ability to hide intracellularly and evade the immune system. It is a slow growing organism that can go dormant. These non growing forms are hard to kill with the currently used antibiotics.

gram positive and gram negative

Bartonella transmission


Bartonella quintana organisms are found in different insects like fleas and ticks but historically it is thought that the main route of transmission is via lice. Lice, like other arthropods spreading these kind of diseases, feed on mammalian blood. When uninfected lice consume blood from an infected person, they become infected with Bartonella.

After swallowing the bacteria, the Bartonella organisms travel to the intestine of the louse and begin to multiply on the surface of the endothelial cells in the intestinal tract. Bartonella stimulates growth of the endothelial layer so that they have a larger surface area to inhabit. Bartonella quintana only replicates every 21 hours. It takes about four days for the bacteria to proliferate in an infected louse. When large numbers are reached, they are released into the feces of the lice for excretion.

After swallowing the bacteria, the Bartonella organisms travel to the intestine of the louse and begin to multiply on the surface of the endothelial cells in the intestinal tract. Bartonella stimulates growth of the endothelial layer so that they have a larger surface area to inhabit. Bartonella quintana only replicates every 21 hours. It takes about four days for the bacteria to proliferate in an infected louse. When large numbers are reached, they are released into the feces of the lice for excretion.

When the lice feeds, skin irritation causes the host to scratch. The skin can break from the scratching and this forces the infected feces into the skin break. Lice can also infect while feeding through the saliva. But the primary route of infection still appears to be through lice feces.

Bartonella quintana is not transmitted to the offspring of infected lice. When lice hetch from the eggs, they feed themselves with the same blood as the previous generation and are thus get infected. Lice lay a large number of eggs – and these eggs are viable for up to two weeks.

While it has long been thought that Bartonella quintana was limited to lice as a vector and humans as a primary host, the progress in laboratory research has shown that we also find this species in monkeys, cats, rodents and dogs. It has been found that all these animals are natural hosts for the organism. This species is much wider spread than previously thought.


Instead of lice, fleas are the main vectors for Bartonella henselae. The process is very similar to that of quintana. Fleas become infected by cats. The fleas spread to other cats and they become infected trough the feces of fleas while cleaning themselves.

Small amounts of flea feces remain on the nails of the cat, and if it scratches a human, the bacteria can penetrate and infect through the skin. Bartonella can also infect the salivary glands of the flea and cat. In this way, transmission can also take place via flea and cat bites.

In fleas, the Bartonella can also infect the reproduction organs. Thus, the infection can be transmitted to the offspring. Fleas are often full of different species of Bartonella, some of which are atypical to their normal host.


A wide variety of ticks, including the Ixodes, Dermacentor, Rhipicephalus and Haemaphysalis can be infected with different species of Bartonella. DNA of Bartonella has been found in between 1 and 60% of the ticks in specific areas. Only one study has confirmed the presence of viable Bartonella by culture. Another problem is that many of the PCR results are not sequenced and the DNA hits can also represent tick symbionts such as Midichloria species.

A recent study suggested vector competence of ticks but Lyme expert Gary P. Wormser has some concerns. He suggests that the presence of a microbial agent within a tick does not imply that the tick might transmit it during the course of blood feeding or that it is pathogenic. The mere presence of Bartonella spp. or their DNA in ticks does not prove vector competence or confer epidemiological significance but it should serve as the impetus to rigorously perform the studies necessary to establish vector competence of ticks for the transmission of Bartonella.

The study suggesting vector competence has got some major limitations. First the system used for infection was artificial and this caused a bacteremia that would rarely be seen in nature. Besides that, the Houston-1 strain used in the experiment may not represent strains found in nature. Primary isolates of naturally occurring strains are extremely fastidious and grow slowly. Whereas the Houston-1 strain readily grows in vitro.


Also biting flies are carriers of Bartonella. Lipoptena, Hippobosca and Meophagus play a role in the cycle of transmission. These different types of flies feed on cows, sheep, horses, deer and so on. The spread from fly to ruminant has not been studied extensively but appears to occur both by saliva and stool.

bartonella transmission

Symptoms of a Bartonella infection

So far 8-12 Bartonella species have been discovered that can cause disease in humans. Their number increases annually. The spectrum of symptoms and clinical manifestations that can be caused by a Bartonella infection is broad.

Other common symptoms include fever, fatigue, muscle ache, joint pain, headache and brainfog. In some cases organs such as the liver and spleen are affected. Bartonella can also infect the eyes, kidneys and lungs. Neurologic disease can include encephalopathy, aseptic meningitis, meningoradiculoneuritis and / or a range of neuro-psychiatric symptoms such as extreme rage, anxiety, panic attacks, depression and psychosis.

The words unusual and rare are regularly used to describe various general symptoms of a case of Bartonellosis. The full spectrum of symptoms is unknown to most physicians.

Eye disorders: Granuloma conjunctivitis, neuroretinitis, sub-retinal lesions, retinitis, intermediary uveitis, chroiditis, retinal vasculitis, optic neuritis

Hearing disorders: Middle ear infection, vertigo, labyrinthitis (both acute and chronic), auditory haullucinations, tinnitus

Neurologic disorders: Seizures (grand mal or otherwise), status epilepticus, epilepsy partialis continua, tremors, aseptic meningitis, meningoencephalitis, cerebral arteritis, peripheral facial nerve palsy, transversal myelitis, radiculitis, Guillain Barré syndrome, polynuropathy, hallucinations (and a broad range of neuro-psychiatric dissorders), cerebral vasculitis, stroke

Skin disorders: Erythema nosodum, petechiae purpura, granuloma dermatitis, angiomatous papillomatosis, granuloma formation in various regions of the skin presenting as painful red nodules that can also occur in the liver or in the spleen

Cardiovascular disorders: Myocarditis, endocarditis (with or without colitis), pericarditis,  cardiac arrhythmia, anemia, heart valve disorders, vasculitis, aneurysms

Bartonella stretch marks

A causal relationship has never been established between the stretch marks and chronic Bartonella infection. These typical stretch marks are associated with a chronic Bartonella infection, especially when there is no overweight or other explanation for the stretch marks they can offer a clue to the treating physician.

Bartonella stretch marks 1

Bartonella stretch marks 1

Bartonella stretch marks 2

Bartonella stretch marks 2

Bartonella stretch marks 3

Bartonella stretch marks 3

Diagnosis of a Bartonella infection

People who come to their doctor with a range of unspecific symptoms are often subjected to a number of tests. These routine tests rarely show anything. White blood cell counts are usually normal or only slightly elevated, platelet count comes back normal and spinal fluid, when examined shows only a mild elevation of leukocytes in the case of neurologic involvement. Liver enzymes are often in the normal range or just a mild elevation. From that point doctors don’t know what is wrong. Sometimes they write a prescription to treat the symptoms in the hope that it works. Rarely a physician thinks about testing for a Bartonella infection in the face of these unspecific symptoms.


Culturing Bartonella organisms from an infected person is a difficult and lengthy process and takes two to six weeks and needs to be done on a specially prepared growth medium. The culture process is often unsuccessful to detect the organisms due to the low number of bacteria in the body. The bacteria are very difficult to isolate, even in people known to be infected.

Culture of lymph nodes, even in obvious cases of cat scratch fever, rarely produces bacteria because the bulk of the swelling consists of a Bartonella-induced immune response and not of the organism itself.


Indirect Fluoressence Assay (IFA) is used to test for antibodies against Bartonella. Nevertheless, this test regularly produces false negatives. False positives can be caused by cross reactions with Epstein-Barr virus, Cytomegalovirus, Coxiella burnetii, Toxoplasma gondii, Chlamydia spp. And Streptococcus pyogenes.

There is a wide variation between strains and between the different species of Bartonella. As a result, there is also large range of antigenic variation. This causes too many problems with this type of test. The test does not respond to all species and strains and cannot differentiate between a passive and active infection. An antibody titer higher than 1:64 is generally considered positive for a recent Bartonella infection.

ELISA is also not very effective. IgG and IgM levels vary greatly; There is no reliable pattern during a Bartonella infection. Edward Breitschwerdt et al report: “It appears that chronic intravascular infection with a Bartonella spp. may induce a degree of immunological anergy, resulting in undetectable levels of organism-specific antibodies in some chronically infected patients.”

Culture + PCR

Most PCR tests can only detect Bartonella if the bacteria are present above a certain level. Because of the tendency of Bartonella to remain at low levels in the body, PCR tests often return negative.

To overcome the obvious problems with test methods for Bartonella A new PCR test was introduced by the company Galaxy Diagnostics. Their test is designed to stimulate the growth of bacteria to detectable PCR levels. The test is 98% effective for a wide range of Bartonella species.

Treatment of Bartonellosis

Up to now no major clinical trials have been performed that demonstrate which antibiotic treatment is most effective for infection with Bartonella. In the case of lymphadenopathy as a result of cat scratch fever, a course of azithromycin appears to improve the patient’s swelling and inconvenience.

In case of endocarditis, surgery is often inevitable. Two weeks of intravenous gentamycin combined with 6 weeks of (oral) doxycycline appears sufficient in these cases. In a case report where surgery was not possible it took 24 months of treatment with doxycycline to achieve a cure. Although Bartonella endocarditis seems rare, it might be unstoppable if it does occur.

In cases where Bartonella has infected the nervous system, the combination of doxycycline with rifampicin seems to yield good results. The dosage may need to be adjusted to achieve good concentrations in the CSF and CNS. Rifampicin might influence the serum levels of doxycycline. The dosage of doxycycline may need to be adjusted accordingly.

Natural resistance is reported for both the macrolides and the fluoroquinolone group of antibiotics. Due to the side effects of fluoroquinolones they should probably be considered as a reserve agent. Chlooremphenicol and fluoroquinolones are, in contrast with other intracellular antibiotics, capable of achieving therapeutic concentrations in red blood cells.

In patients with peliosis hepatitis and bacillary angiomatosis, an extended course of antibiotics of 3-4 months with macrolide or tetracycline antibiotics seems to be effective.

Treatment failure

Treatment of complicated cases will have to be adjusted on an individual basis. Bartonella can become rapidly resistant to antibiotics. Relapse is widely reported and treatment with two intracellular acting antibiotics produces consistently better results than mono therapy and can possibly prevent the development of resistance.

The reason for therapy failure is not yet fully understood but persistence in red blood cells, bone marrow and endothelial cells and difficult to reach places is suggested. Also, most antibiotics act only as bacteriostatics against Bartonella. The lack of bactericidal activity of most antibiotics against Bartonella is suggested as a possible reason why patients relapse.

Bartonella are hard to kill with some antibiotics. Bacteria that don’t grow or grow slowly can survive short courses of many antibiotics since most antibiotics only work against active replicating bacteria.

This might be another reason why rifampicin can be more successful against Bartonella because rifampicin is one of the few drugs that has some activity against dormant, non growing bacteria. Rifampicin has been shown to be moderately bactericidal in some studies.

Rifamycins work by peak concentration and this is why using one large dose might be more effective compared to two small doses daily.

Diagnosis and treatment can be difficult. We do not know how to treat this infection in a subset of patients with a chronic presentation. In these cases antibiotic treatment of two or three antibiotics (doxycycline, azithromycine/clarithromycine and rifampicin with or without hydroxychloroquine) for one to two years has been used with success. See references for these case reports.

Other resources