Trench Fever (Bartonella quintana infection): A Comprehensive Disease Characteristics Report

Disease: Trench Fever · MONDO ID: MONDO:0005991 · Category: Infectious Disease Causal agent: Bartonella quintana (α-proteobacteria; formerly Rochalimaea quintana) ICD-10: A79.0 · ICD-11: 1C30.0 · MeSH: D014205 (Trench Fever) Evidence base: aggregated disease-level literature (reviews, case-control studies, one randomized trial, experimental microbiology). No individual EHR or genomic data files were provided; this report synthesizes primary literature retrieved via PubMed (27 papers reviewed, 11 findings confirmed).

Summary

Trench fever is a louse-borne bacterial infection caused by Bartonella quintana, a small, fastidious, facultatively intracellular α-proteobacterium that is restricted to a single natural reservoir — humans — and a single principal vector, the human body louse (Pediculus humanus humanus). First characterized in 1915 among soldiers in the trenches of World War I (hence the name), the disease has re-emerged over the past three decades as "urban trench fever" among homeless and socially marginalized populations in high-income cities in the US and Europe (PMID: 16494745). It is fundamentally a disease of poverty and body-louse infestation, not a genetic disease; there are no known human causal genes, susceptibility loci, or heritable risk factors.

Clinically, B. quintana produces a broad spectrum running from the classic self-limited relapsing febrile illness ("five-day fever," with headache and severe shin/tibial bone pain) through chronic afebrile bacteremia, blood-culture-negative endocarditis, chronic lymphadenopathy, and — particularly in immunocompromised hosts such as people with HIV — the vasoproliferative disorders bacillary angiomatosis and peliosis (PMID: 16494745; PMID: 15891120). Two mechanistic axes dominate the pathophysiology: (1) the laterally acquired Trw type IV secretion system (T4SS) mediates host-specific adhesion to and invasion of erythrocytes, producing the hallmark long-lasting intraerythrocytic bacteremia; and (2) a suite of factors — the Vomps trimeric autotransporter adhesins, the VirB/VirD4 T4SS with its Bep effectors, and the BafA autotransporter (a VEGF analog) — converge on HIF-1/VEGF signaling to drive endothelial proliferation, tube formation, and angiogenesis (PMID: 20548954; PMID: 32678094; PMID: 21536788; PMID: 23163798).

Diagnosis rests on serology, blood culture (often prolonged), and molecular detection (PCR of gltA/ftsZ/16S rRNA). The best-supported treatment for chronic bacteremia is doxycycline plus gentamicin, validated by a randomized open trial (eradication in 7/7 per-protocol treated vs 2/9 controls, P=0.003) (PMID: 12821469) and a systematic review/meta-analysis (PMID: 23602630). Prevention is fundamentally body-louse control and improved hygiene/social conditions; there is no vaccine. This report synthesizes 11 confirmed findings across all 15 template sections.

Key Findings

Finding 1 — Trench fever is caused by Bartonella quintana, a human-restricted, louse-borne pathogen

Bartonella quintana is an α-proteobacterium (formerly Rochalimaea quintana) whose lifestyle is restricted to human hosts and louse vectors. It was the first Bartonella species definitively linked to human disease and was characterized as the agent of trench fever, a disease "described in 1915 on the basis of natural and experimental infections in soldiers" (PMID: 16494745). After decades of relative obscurity, it re-emerged as a pathogen of the urban homeless: in a study of homeless persons, B. quintana was isolated from blood culture in 5.3% of patients (PMID: 15643300).

The verbatim evidence: "Bartonella quintana, a pathogen that is restricted to human hosts and louse vectors, was first characterized as the agent of trench fever. The disease was described in 1915 on the basis of natural and experimental infections in soldiers" (PMID: 16494745).

Key identifiers: MONDO:0005991; ICD-10 A79.0; ICD-11 1C30.0; MeSH D014205. This is an infectious, aggregated disease-level entity — knowledge is derived from case series, cohort studies, microbiology, and molecular biology, not from Mendelian/EHR genetic resources (no OMIM/Orphanet gene entry applies because it is not heritable). Common synonyms: five-day fever, quintan fever, His-Werner disease, shin bone fever, wolhynia fever, urban trench fever.

Finding 2 — B. quintana causes a clinical spectrum from classic trench fever to chronic bacteremia, endocarditis, and bacillary angiomatosis

B. quintana is "responsible for a wide spectrum of conditions, including chronic bacteremia, endocarditis, and bacillary angiomatosis" (PMID: 16494745). The classic acute syndrome is a relapsing fever recurring at ~5-day intervals ("quintana" = fifth), with headache, dizziness, and characteristic severe pain in the shins/tibiae. Beyond the acute illness, chronic afebrile bacteremia can persist for months; endocarditis is an important severe outcome. In a series from Sfax, Tunisia, B. quintana endocarditis "represents 9.8% of all endocarditis" in that setting, with high IgG titers (≥1:800) supporting the diagnosis (PMID: 15891120). Chronic infection can also present as isolated lymphadenopathy/adenomegaly, including in seronegative patients (PMID: 8727894). In immunocompromised hosts (notably HIV), the organism drives vasoproliferative disease — bacillary angiomatosis and peliosis.

Finding 3 — Trw T4SS mediates erythrocyte invasion; BafA drives VEGF-dependent angiogenesis

Two experimentally demonstrated virulence mechanisms anchor the pathophysiology.

(1) Erythrocyte invasion via the Trw T4SS. Using signature-tagged mutagenesis in surrogate Bartonella species, the Trw type IV secretion system was shown to be directly involved in erythrocyte adhesion; its genes "encode components of the type IV secretion system (T4SS) Trw, demonstrating that this virulence factor laterally acquired by the Bartonella lineage is directly involved in adherence to erythrocytes" (PMID: 20548954). This host-specific adhesion enables the hallmark long-lasting intraerythrocytic bacteremia.

(2) VEGF-driven vasoproliferation via BafA. The Bartonella autotransporter BafA acts as a VEGF analog: "BafA interacts with vascular endothelial growth factor (VEGF) receptor-2 and activates the downstream signaling pathway, suggesting that BafA functions as a VEGF analog. A BafA homolog from a related pathogen, Bartonella quintana, is also functional" (PMID: 32678094). This explains the vasoproliferative lesions (bacillary angiomatosis) directly at the molecular level.

Findings 4 & 10 — Treatment: doxycycline + gentamicin eradicates chronic bacteremia (randomized + meta-analytic evidence)

The best-supported regimen for chronic B. quintana bacteremia is doxycycline plus gentamicin. A randomized open trial (Foucault, Raoult, Brouqui 2003) in homeless patients with blood-culture-positive B. quintana compared gentamicin 3 mg/kg/day IV for 14 days + doxycycline 200 mg/day orally for 28 days versus no treatment. "Intention-to-treat analysis of 20 included patients showed eradication of bacteremia in 7 out of 9 treated patients versus 2 out of 11 untreated controls (P = 0.01). In the per-protocol analysis, eradication was obtained for 7 out of 7 treated patients versus 2 out of 9 untreated controls (P = 0.003)" (PMID: 12821469).

A systematic review and meta-analysis of human bartonellosis treatment corroborates this: "In chronic bacteremia, gentamicin and doxycycline significantly increased the resolution rate" (PMID: 23602630) — though the authors note the evidence base is limited and dominated by observational studies at high risk of bias. Expert-guideline regimens (Rolain et al. 2004) extend this: endocarditis typically requires prolonged therapy (often doxycycline + an aminoglycoside) and may require valve surgery; bacillary angiomatosis responds to macrolides (erythromycin) or doxycycline given for months. NCIT term suggestions: Doxycycline (NCIT:C692), Gentamicin (NCIT:C575), Erythromycin (NCIT:C608).

Finding 5 — B. quintana has a reduced ~1.58 Mb genome shaped by secondary genome reduction; T4SS drives host adaptation

Comparative genomics of the genus shows Bartonella are α-proteobacteria with host-restricted lifestyles characterized by long-lasting intraerythrocytic infection in a specific mammalian reservoir and arthropod transmission. Among sequenced genomes (circular chromosomes 1.44–2.62 Mb; 1,283–2,136 genes; a synthenic core of 959 genes), B. quintana underwent "massive secondary genome reduction" — it has the smallest genome (~1.58 Mb, ~1,300 genes) — consistent with adaptation to a single human host plus louse vector (PMID: 19696500). The type IV secretion systems (Trw, VirB/VirD4) were laterally acquired and are "type IV secretion systems that adopted prominent roles in host adaptation and specificity" (PMID: 19696500). This reductive evolution reflects the pathogen's dependence on host-derived metabolites (host-integrated metabolism) and its narrow ecological niche.

Finding 6 — Bacillary angiomatosis is epidemiologically linked to homelessness, low income, and lice, with subcutaneous and lytic bone lesions

A case-control study of 49 HIV-associated bacillary angiomatosis-peliosis patients versus 96 matched controls found that 23/49 (47%) were B. quintana and 26/49 (53%) B. henselae. Crucially, tissue tropism differed by species: "Subcutaneous and lytic bone lesions were strongly associated with B. quintana, whereas peliosis hepatis was associated exclusively with B. henselae" (PMID: 9407154). Risk factors also segregated: patients "with B. quintana were clustered and were characterized by low income (P=0.003), homelessness (P = 0.004), and exposure to lice (P= 0.03)", whereas B. henselae was tied to cat/flea exposure (PMID: 9407154). This confirms both the skin/bone tissue tropism of B. quintana and its socioeconomic risk profile.

Finding 7 — Transmission is via infectious body-louse feces (horizontal); no true transovarial (vertical) transmission

Humans acquire B. quintana through contaminated body-louse feces, not the louse bite directly: "Horizontal transmission from the body louse vector (Pediculus humanus humanus) to a human host occurs through contact with infectious louse feces containing a high concentration of the bacteria" — typically inoculated into skin abrasions or bite-scratch excoriations (PMID: 35803580). Controlled experiments testing vertical transmission found bacterial DNA on egg surfaces (fecal contamination of the egg sheath) but no viable organisms internally: "viable B. quintana could not be cultured from the hemolymph of adult female lice or from within eggs that were surface sterilized, indicating a lack of true transovarial transmission" (PMID: 35803580). Vertical transfer therefore "probably has a limited impact on the dynamics of transmission to humans." This has direct public-health implications: interrupting the human↔louse cycle (delousing) breaks transmission.

Finding 8 — Vomps adhesins and the VirB/D4 T4SS–Bep effector system converge on HIF-1/VEGF to drive vasoproliferation

B. quintana expresses variably expressed outer membrane proteins (Vomps), which are trimeric autotransporter adhesins (TAAs) analogous to B. henselae's Bartonella adhesin A (BadA). In a study of adherence under static and bloodstream-like dynamic flow, investigators "analyzed three different TAAs (Bartonella adhesin A [BadA] of Bartonella henselae, variably expressed outer membrane proteins [Vomps] of Bartonella quintana, and Yersinia adhesin A [YadA] of Yersinia enterocolitica) for mediating bacterial adherence to ECM and endothelial cells" (PMID: 21536788) — establishing Vomps as B. quintana's ECM/endothelial adhesins (binding fibronectin, collagen).

Mechanistically, in the closely related B. henselae the pathway is well defined: "Expression of Bartonella adhesin A (BadA) is crucial for bacterial autoagglutination, adhesion to host cells, binding to extracellular matrix proteins and proangiogenic reprogramming via activation of hypoxia inducible factor (HIF)-1" (PMID: 18627378). In parallel, the VirB/VirD4 T4SS delivers effector proteins into endothelial cells: "VirB/D4 translocates several Bartonella effector proteins (Beps) into the cytoplasm of infected ECs, resulting, e.g. in uptake of bacterial aggregates via the invasome structure, inhibition of apoptosis and activation of a proangiogenic phenotype" (PMID: 23163798). Notably, BadA and VirB/D4 are frequently mutually exclusive during in vitro passage — a phase-variation phenomenon relevant to culture-dependent diagnostics.

Finding 9 — Humans are the primary reservoir, but B. quintana DNA is increasingly detected in animals; dogs model Bartonella endocarditis

The natural reservoir of B. quintana is humans (NCBI Taxon 9606), transmitted by the body louse. However, molecular surveys increasingly detect B. quintana DNA in non-human hosts. A survey of pet cats in Urmia, Iran found "15 % of the cats (30 out of 200 blood samples) tested positive for the B. quintana gene, with a 95 % confidence interval of 10.71 % to 20.61 %" (100% sequence identity to the reference) (PMID: 38199070). For comparative disease modeling, dogs develop naturally occurring Bartonella endocarditis, making "canids... the most interesting naturally occurring animal model for the human disease" (PMID: 12860639). Experimental mechanistic work uses surrogate rodent-adapted species — "mouse-specific Bartonella birtlesii, human-specific Bartonella quintana, cat-specific Bartonella henselae and rat-specific Bartonella tribocorum" in adhesion/invasion assays (PMID: 20548954) — and the body louse itself serves as an in-vivo vector model.

Finding 11 — Epidemiology: ~5% of sheltered homeless bacteremic; rising; worldwide wherever body lice occur

In a 4-year study of 930 homeless people in Marseilles, lice were found in 22%, and B. quintana was isolated from blood culture in 50 (5.3%); critically, "the number of bacteremic patient increased from 3.4% to 8.4% (p = 0.02) over the 4 years of the study" — evidence of a rising trend (PMID: 15643300). Geographically, the pathogen is cosmopolitan wherever body-louse infestation occurs: a molecular survey of body lice detected "the presence of B. quintana in lice collected from all locations except the Congo" — spanning France, Russia, Peru, Zimbabwe, and Burundi (PMID: 9986818). B. quintana is one of only three body-louse-borne human pathogens, alongside Rickettsia prowazekii (epidemic typhus) and Borrelia recurrentis (louse-borne relapsing fever); co-exposure occurs, with documented R. prowazekii seroconversion in a B. quintana-bacteremic homeless person (PMID: 15891141).

Mechanistic Model / Interpretation

Ordered causal chain: from louse feces to clinical disease

1.  Body louse (Pediculus humanus humanus) feeds on a bacteremic human
        │ leads to
2.  B. quintana replicates in the louse gut and is shed in LOUSE FECES
    (high bacterial concentration; NOT transovarial — no vertical spread)
        │ results in
3.  Contaminated feces inoculated into skin abrasions / bite-scratch
    excoriations of a new human host (horizontal transmission)
        │ leads to
4.  Bacteria seed a primary niche (endothelial cells / dermal tissue)
        │ then (via Trw T4SS)
5.  Trw type IV secretion system mediates HOST-SPECIFIC ADHESION to and
    INVASION of erythrocytes
        │ results in
6.  Long-lasting INTRAERYTHROCYTIC BACTEREMIA (immune-privileged niche;
    enables relapsing fever + chronic bacteremia + louse re-acquisition)
        │
        ├─► BRANCH A (acute/chronic systemic disease):
        │       relapsing 5-day fever, headache, severe shin/bone pain;
        │       chronic afebrile bacteremia → endocarditis (valve
        │       vegetations, culture-negative)
        │
        └─► BRANCH B (vasoproliferative disease, esp. immunocompromised):
                Vomps (TAAs) adhere to ECM (fibronectin/collagen) + endothelium
                        │ + VirB/D4 T4SS translocates Bep effectors
                        │ + BafA autotransporter binds VEGFR-2 (VEGF analog)
                        │ results in
                HIF-1 activation → VEGF secretion (autocrine/paracrine loop)
                inhibition of endothelial apoptosis; invasome-mediated uptake
                        │ leads to
                ENDOTHELIAL PROLIFERATION → tube formation → ANGIOGENESIS
                        │ results in
                Bacillary angiomatosis (skin, LYTIC BONE lesions),
                vasoproliferative tumors

Upstream vs downstream. The initiating lesion is environmental/behavioral (louse infestation + poverty), not genetic. The most upstream molecular event in host colonization is Trw-mediated erythrocyte tropism, which sustains the bacteremic reservoir. The vasoproliferative arm is downstream of endothelial adhesion (Vomps/BadA) and effector delivery (VirB/D4-Bep, BafA), all converging on the HIF-1 → VEGF hub. Note that Trw (erythrocyte invasion) and VirB/D4 (endothelial reprogramming) are distinct T4SSs with distinct target cells. The HIF-1/VEGF steps are directly demonstrated for B. henselae BadA and inferred for B. quintana via its homologous Vomps and functional BafA homolog.

Cell types and processes (ontology suggestions). - Erythrocyte (CL:0000232) — intracellular niche; GO:0007155 cell adhesion; GO:0044409 entry into host. - Endothelial cell (CL:0000115) — angiogenesis target; GO:0001525 angiogenesis; GO:0043066 negative regulation of apoptotic process; GO:0001666 response to hypoxia (HIF-1). - Anatomical sites (UBERON): blood (UBERON:0000178), skin/dermis (UBERON:0002067), bone (UBERON:0001474), heart valve/endocardium (UBERON:0002165), lymph node (UBERON:0000029), liver (peliosis, mainly B. henselae; UBERON:0002107). - Chemical/biomarker entities (CHEBI): doxycycline (CHEBI:50845); gentamicin (CHEBI:27412).

Comparative species pathophysiology

Section-by-Section Template Coverage

• §1 Disease Info / §5 Infectious agent: B. quintana (α-proteobacterium); MONDO:0005991, ICD-10 A79.0, ICD-11 1C30.0, MeSH D014205. Synonyms: five-day/quintan/shin-bone/wolhynia fever, His-Werner disease, urban trench fever. Aggregated disease-level knowledge.
• §2 Etiology / risk factors: Infectious etiology; risk = homelessness, poverty, low income, body-louse infestation, poor hygiene, cold climate/crowding, alcoholism; immunosuppression (HIV) for vasoproliferative disease. No genetic causal or protective factors (not heritable); no meaningful gene-environment interaction reported.
• §3 Phenotypes: Fever HP:0001945 (relapsing, ~5-day periodicity), Headache HP:0002315, Bone pain HP:0002653 (shin/tibia), Splenomegaly HP:0001744, Bacteremia HP:0031864, Endocarditis HP:0100584, Lymphadenopathy HP:0002716. Onset acute (adult, occupational/social exposure); severity variable (mild self-limited → severe endocarditis).
• §4 Genetic/molecular: Not applicable to the human host — no causal genes, pathogenic variants, modifier genes, or chromosomal abnormalities. Relevant "genes" are bacterial virulence loci: trw (T4SS), virB/virD4 (T4SS + bep effectors), vomp family, bafA.
• §6 Mechanism: See causal chain above.
• §7 Anatomy: Blood/erythrocytes (primary), vascular endothelium, skin/dermis, bone (lytic lesions), cardiac valves/endocardium, lymph nodes, spleen.
• §8 Temporal: Incubation ~15–30 days; acute relapsing fever often self-limited over weeks; chronic bacteremia/endocarditis can persist months–years if untreated (chronic, insidious course).
• §9 Epidemiology: ~5% bacteremia prevalence in sheltered homeless (rising); worldwide wherever body lice occur; predominantly adult; male predominance in homeless cohorts (demographic, not biological). Not inherited — inheritance/penetrance/expressivity sections not applicable.
• §10 Diagnostics: Serology (IFA IgG, titers ≥1:800 support endocarditis), prolonged blood culture, PCR/sequencing of gltA, ftsZ, 16S rRNA; histopathology (Warthin-Starry silver stain) for bacillary angiomatosis; echocardiography for endocarditis. Differential: culture-negative endocarditis (Coxiella, Tropheryma), other Bartonella, epidemic typhus, relapsing fever.
• §11 Prognosis: Acute trench fever generally self-limited and non-fatal; endocarditis carries substantial morbidity/mortality and often needs valve surgery; good response to appropriate antibiotics.
• §12 Treatment: Doxycycline + gentamicin (chronic bacteremia, RCT-supported); doxycycline ± aminoglycoside for endocarditis (+ surgery); macrolides/doxycycline for bacillary angiomatosis.
• §13 Prevention: Body-louse control (delousing, hygiene, insecticide-treated clothing), improved housing/social conditions, health education; no vaccine (PMID: 20822446).
• §14 Other species: Human reservoir (NCBI:txid9606); molecular detection in cats/dogs/other animals; dogs = natural endocarditis model.
• §15 Model organisms: Surrogate rodent-adapted Bartonella (B. birtlesii, B. tribocorum); body louse as vector model; endothelial-cell and erythrocyte in-vitro assays.

Evidence Base

Limitations and Knowledge Gaps

1. Mechanistic inference across species. Much of the vasoproliferative signaling detail (HIF-1 activation, VirB/D4-Bep effector biology, BadA structure–function) is demonstrated in B. henselae and inferred for B. quintana via the homologous Vomps/BafA systems. Direct B. quintana endothelial-reprogramming studies are comparatively sparse. The BafA VEGFR-2 interaction was shown functional for the B. quintana homolog, but the full downstream cascade in B. quintana is extrapolated.

2. Treatment evidence is thin. The pivotal RCT enrolled only 20 patients (PMID: 12821469), and the meta-analysis explicitly flags that most treatment evidence is observational and at high risk of bias (PMID: 23602630). Optimal duration for endocarditis, and management of relapse, remain guideline/expert-opinion driven.

3. Epidemiology is likely underestimated. Fastidious culture requirements and non-specific clinical presentation mean cases are under-ascertained; prevalence figures derive largely from urban homeless cohorts in France and may not generalize globally. Population-level incidence rates (new cases/100,000/yr) are not well established.

4. Expanding host range uncertain in significance. Detection of B. quintana DNA in cats, dogs, non-human primates, and other animals (PMID: 38199070) raises the question of whether animals are competent reservoirs or incidental/spillover hosts. Whether these represent transmission risk to humans is unresolved.

5. No human genetic component — the template's genetic/variant/inheritance/heritable-model sections are not applicable; any "susceptibility" is socioeconomic and behavioral, not genomic.

6. Quality-of-life and disability data specific to trench fever (EQ-5D, SF-36, PROMIS) are essentially absent in the literature; QoL impact is inferred from the burden of chronic bacteremia/endocarditis and the comorbidities of the affected homeless population.

Proposed Follow-up Experiments / Actions

1. Direct B. quintana vasoproliferation assays. Test purified B. quintana BafA and Vomps in human endothelial cells for HIF-1 stabilization, VEGF secretion, tube formation, and apoptosis inhibition — to replace B. henselae inference with species-specific data.

2. Larger, multicenter treatment trial. A pragmatic RCT (or well-designed prospective registry) across multiple homeless-service settings to define the optimal antibiotic regimen and duration for chronic bacteremia and endocarditis, and to quantify relapse.

3. Global surveillance via body lice. Extend the molecular-surveillance approach (PMID: 9986818) to systematically map B. quintana prevalence in body lice and homeless populations worldwide, including incidence estimation.

4. Reservoir-competence studies in animals. Determine whether cats/dogs/primates with detectable B. quintana DNA harbor viable, transmissible bacteria and whether alternative arthropod vectors participate — clarifying zoonotic risk.

5. Public-health intervention evaluation. Rigorously evaluate delousing/hygiene/housing interventions for their effect on B. quintana incidence, integrating vector control per PMID: 20822446.

6. Biomarker/diagnostic development. Develop rapid point-of-care serologic or molecular assays suited to low-resource, homeless-service settings to improve case ascertainment and shorten time-to-treatment.

Consensus Answer

Trench fever is a louse-borne bacterial infection caused by Bartonella quintana (MONDO:0005991; ICD-10 A79.0), a human-restricted α-proteobacterium transmitted through the feces of the body louse (Pediculus humanus humanus) and re-emerging among the urban homeless worldwide. It is not a genetic disease — risk is socioeconomic (homelessness, poverty, body lice, poor hygiene; immunosuppression for vasoproliferative forms) — and it spans a clinical spectrum from classic relapsing five-day fever with shin pain to chronic bacteremia, culture-negative endocarditis, and bacillary angiomatosis, driven mechanistically by Trw-T4SS–mediated erythrocyte invasion plus Vomps/BadA-, VirB/D4-Bep–, and BafA-driven HIF-1/VEGF endothelial proliferation. It is effectively treated with doxycycline plus gentamicin (randomized-trial evidence) and prevented by body-louse control; no vaccine exists.