key: cord-1048667-4r2kcxyu
authors: McCormick, David W; Rowan, Sarah E; Pappert, Ryan; Yockey, Brook; Dietrich, Elizabeth A; Petersen, Jeannine M; Hinckley, Alison F; Marx, Grace E
title: Bartonella Seroreactivity among Persons Experiencing Homelessness During an Outbreak of Bartonella quintana in Denver, Colorado, 2020
date: 2021-05-08
journal: Open Forum Infect Dis
DOI: 10.1093/ofid/ofab230
sha: 152033b6f87af88f3fb9cbb28a78a87b10666c9b
doc_id: 1048667
cord_uid: 4r2kcxyu

During a recent outbreak of Bartonella quintana disease in Denver, 15% of 241 persons experiencing homelessness who presented for SARS-CoV-2 testing were seroreactive for Bartonella. Improved recognition of B. quintana disease and prevention of louse infestation are critical for this vulnerable population.

The louse-borne bacterium Bartonella quintana is the most frequent cause of vectorborne disease among persons experiencing homelessness (PEH) in the United States and Europe [1] . Bartonella quintana disease is characterized by a wide range of clinical presentations, including a distinctive febrile illness and culture-negative endocarditis; many individuals may have asymptomatic or mildly symptomatic infections [2] . Serology can assist with the diagnosis of B. quintana infection, but Bartonella seroreactivity does not necessarily indicate active disease and does not differentiate between Bartonella species [3] . Individuals infected with B. quintana can remain seroreactive for years, even following effective treatment [3] .

During the summer of 2020, an outbreak of B. quintana disease was identified among PEH in the Denver, Colorado metropolitan area [4] . To estimate the prevalence of past exposure to B. quintana among this population, we evaluated seroreactivity to Bartonella among PEH in Denver using residual serum samples obtained for SARS-CoV-2 serology testing. Statistical Methods: We described continuous variables using median and interquartile ranges (IQR) and categorial data using counts and percentages. Bartonella seroreactivity was defined as a titer ≥1:512. We used the Mann-Whitney U test to compare age between seroreactive and non-seroreactive persons and the χ 2 test or Fisher Exact test to examine A c c e p t e d M a n u s c r i p t 5 associations between categorical variables and seroreactivity status. All statistical tests were two-sided and we considered a p-value of ≤0.05 statistically significant. Statistical calculations were performed using R Version 4.0.3 (R Foundation for Statistical Computing) [5] .

We used logistic regression to examine the strength of association between available demographic variables and seroreactivity to Bartonella. We included self-identified gender, self-identified race/ethnicity, age, collection setting, and SARS-CoV-2 serostatus in a multivariable model.

To examine for an effect due to cross-reactivity between SARS-CoV-2 and Bartonella antibodies, we repeated the analyses after censoring all PEH who tested positive for SARS-CoV-2.

Residual serum samples were available from 241 participants at encampments Persons who were seroreactive to Bartonella were significantly older (median age 50.5 years, IQR 40-57 years) than persons who were not seroreactive (median age 43 years, IQR 34-54 years, p=0.04, Table 1 ). Seroreactive persons had similar gender and race/ethnicity distributions compared to those who were not seroreactive. In the adjusted logistic regression model, gender, race/ethnicity, collection setting, or age were not A c c e p t e d M a n u s c r i p t 6 significantly associated with Bartonella seroreactivity (Supplemental Table 1) . After censoring persons who tested positive for SARS-CoV-2, a similar proportion were Bartonella seroreactive (Supplemental Table 2 ).

We identified a high proportion of Bartonella seroreactivity among PEH in Denver, Colorado during June-July 2020 in the context of a recently recognized outbreak of B.

quintana disease. These findings in Denver are consistent with prior studies in urban settings and indicate that B. quintana disease remains a concern in PEH in the US. In 1996, an outbreak of B. quintana endocarditis among PEH in Seattle prompted a serologic study of PEH presenting for clinical care at a single clinic, 20% were seroreactive compared to 2% of blood donor controls [6] ; another study reported that 9.5% of a convenience sample of persons who sought care at a free clinic in Los Angeles in 2002 were seroreactive [7] . Studies examining seroreactivity to Bartonella in the United States among persons who inject drugs found seroreactivity to B. quintana antigens in 10% of individuals in Baltimore (1996) [8] and 2% in New York (2001) [9] . During 2012-2014, there were 6 confirmed cases of B.

quintana endocarditis among PEH in Alaska. Body lice positive for B. quintana were collected from two of these patients [10] . More recent data on seroprevalence in the United States is lacking.

It is difficult to directly compare our results to other studies given different serological assays/test reagents and subjectivity in determining titers [11, 12] . The ≥1:512 titer used to define Bartonella seroreactivity was chosen for this serosurvey to optimize specificity. While this high titer threshold may have resulted in misclassifying some persons with very recent Bartonella infection as non-seroreactive, it allows for high confidence that persons classified as seroreactive in this study truly represent those with prior or current A c c e p t e d M a n u s c r i p t 7 Bartonella infection. Prior studies suggest that high titers are present in active, recent, or recurrent infection [11, 12] .

Older age was associated with seroreactivity in this study, which may reflect higher prevalence of risk factors for Bartonella infection. Risk factors associated with Bartonella seroreactivity in previous studies included alcohol abuse, tobacco abuse, intravenous drug use, and homelessness [6] . Body lice infestation is a well-recognized risk factor for B.

quintana infection among PEH [2, 13] , and body lice infestation among PEH in San Francisco was associated with sleeping outdoors, male gender, and Black, non-Hispanic race/ethnicity [14] .

This study is subject to at least four limitations. First, the presence of serum IgG does not distinguish between current and past infections; thus, these results in the absence of clinical symptoms, epidemiologic data, or other laboratory evidence are insufficient to confirm recent infections or a common source of B. quintana infection among these participants. The Bartonella IFA is not specific to B. quintana and cross-reacts with antibodies to other Bartonella species. Thus, the true rate of seroreactivity due to infection with B. quintana among PEH in Denver may be lower than presented here. Second, our findings might not be representative of PEH in Denver or in other urban areas, given the risk for selection bias due to use of residual specimens from a convenience sampling strategy.

Third, health and behavioral risk factor data were not collected, making it impossible to evaluate critical determinants of health, such as access to hygienic services and behavioral and medical healthcare services. Fourth, we did not have information regarding shelter or encampments where participants may have visited or slept. This information could provide useful insights as to common sources of lice exposure or settings where access to hygiene and lice mitigation are more challenging.

A c c e p t e d M a n u s c r i p t 8 The high proportion of seroreactive participants in this report suggests that B.

quintana infection is of concern among PEH in Denver. Overcrowded living conditions and limited access to hygienic services for people without stable housing are likely to continue to drive this disease of poverty [2] . Active surveillance and treatment of body lice infestation, especially among at-risk individuals and in communities where B. quintana infections or outbreaks are detected, can be implemented to prevent infection. Clinicians should be vigilant for symptoms among PEH that might suggest B. quintana disease such as non-specific febrile syndromes or symptoms of endocarditis and consider sending clinical specimens for B. quintana molecular diagnostic testing. If culture is ordered, the microbiology laboratory should be notified that B. quintana infection is suspected in order to optimize culture techniques, including extending the incubation period for ≥21 days [3] . These results underscore the need for heightened clinical awareness of B. quintana infection and improved access to hygienic services in this population given louse-borne transmission of this bacterium [2] .

A c c e p t e d M a n u s c r i p t 9

Zoonotic and Vector-Borne Infections among Urban Homeless and Marginalized People in the United States and Europe

Infections in the homeless

Bartonella Species, an Emerging Cause of Blood-Culture-Negative Endocarditis

Rare trench fever found among Denver's homeless population

Seroprevalence to Bartonella quintana among patients at a community clinic in downtown Seattle

Prevalence study of antibody to ratborne pathogens and other agents among patients using a free clinic in downtown Los Angeles

Antibodies to Bartonella species in inner-city intravenous drug users in Baltimore

Evidence of rodent-associated Bartonella and Rickettsia infections among intravenous drug users from central and east harlem

Bartonella quintana Endocarditis Following Body Louse Exposure

Bartonella-associated infections

Value of microimmunofluorescence for diagnosis and follow-up of Bartonella endocarditis

Survey of the seroprevalence of Bartonella quintana in homeless people

Risk factors for human lice and bartonellosis among the homeless

We thank the clients and staff at homeless shelters and encampments in Denver for their assistance and participation. We thank Jesse Chavez, Charles Chen, Tracy Scott, and Jose Silva for their assistance with sample collection. We thank Laura Triplett, Julia Frey, Rosie Horst, and Stephanie Sanders at Denver Health for their assistance processing samples. 

A c c e p t e d M a n u s c r i p t 10 A c c e p t e d M a n u s c r i p t 11 A c c e p t e d M a n u s c r i p t 12