Bartonella quintana in Body Lice from Scalp Hair of Homeless Persons, France

To the Editor: Bartonella quintana is a body louse–borne human pathogen that can cause trench fever, bacillary angiomatosis, endocarditis, chronic bacteremia, and chronic lymphadenopathy (1). Recently, B. quintana DNA was detected in lice collected from the heads of poor and homeless persons from the United States, Nepal, Senegal, Ethiopia, and the Democratic Republic of the Congo and in nits in France (2,3). The head louse, Pediculus humanus capitis, and the body louse, Pediculus humanus humanus, are obligatory ectoparasites that feed exclusively on human blood (4). Outside of their habitats, the 2 ecotypes are morphologically indistinguishable (1). Sequence variation in the PHUM540560 gene discriminates between head and body lice by determining the genotype of the lice (5). While surveying for trench fever among homeless persons in shelters in Marseille, France during October 2012–March 2013, we investigated the presence of B. quintana DNA in nits, larvae, and adult lice collected from mono-infested and dually infested persons and determined the genotypes of the specimens. The persons included in this study received long-lasting insecticide-treated underwear; lice were collected by removing them from clothing, including underwear, pants, and shirts. Because body lice reside in the clothing of infested persons except when feeding, they are sometimes called clothing lice. A total of 989 specimens were tested, including 149 (83 from clothing and 66 from hair) first–instar larvae hatched in the laboratory from eggs collected from 7 dually infested persons, and 840 adult body lice collected from the clothing of 80 monoinfested patients. We included DNA isolated from 3 nits collected from the hair of a mono-infested person who had previously been confirmed as positive for B. quintana 6) (Table). Total DNA was extracted by using an EZ1 automated extractor (QIAGEN, Courtaboeuf, France) and subjected twice to real-time PCR specific for B. quintana. The first PCR targeted the 16S-23S intergenic spacer region. Positive samples were confirmed by using a second real-time PCR targeting the yopP gene (6). Samples that tested positive for B. quintana DNA were analyzed by multiplex real-time PCR that targeted the PHUM540560 gene (5). We used head and body lice that had known genotypes positive controls. Negative controls were included in each assay. Of the hatched larvae, 5 (6%) of the 83 recovered from clothing and 7 (11%) of 66 from the hair (Table) of 4 of the 7 dually infested persons were positive for B. quintana DNA (online Technical Appendix Table 1 wwwnc.cdc.gov/EID/article/20/5/131242-Techapp1.pdf). Of the 840 adult body lice, 174 (21%) collected from 42 (53%) of 80 of the mono-infested persons contained B. quintana DNA (Table, online Technical Appendix 2). The multiplex real-time PCR that targeted the PHUM540560 gene clearly identified all nits, larvae, and adult lice as belonging to the body lice lineage. Negative controls remained negative in all PCR-based experiments. For 2 decades, B. quintana DNA has been regularly detected in lice collected from the heads of persons living in poverty, but it had not been detected in head lice that infest schoolchildren (7,8). All of the lice collected during this study that tested positive for B. quintana from homeless persons were body lice, including some that were recovered from hair. This observation supports our assertion that body lice are not confined to the body. The 3 eggs that were removed from the hair of a mono-infested homeless person whose samples tested positive for B. quintana were also body lice. During the clinical examination, no adult head lice or adult body lice were found on that person, confirming that the patient had been heavily infested with body lice in the past, not head lice. The nits were most likely laid by body lice that migrated toward the patient’s head. When a member of