During March 2014–June 2017, we screened 2,236 febrile patients seeking treatment at The People’s Liberation Army 990 Hospital in Xinyang, Henan Province, eastern central China (Appendix Figure) with a history of tick bite or field activity within the previous month for infection with R. japonica. We extracted DNA from peripheral blood samples collected during the acute phase of illness and screened them by a nested PCR concurrently targeting the citrate synthase gene (gltA), 17-kDa antigen–encoding gene (htrA), and outer membrane protein A–encoding gene (ompA), as previously described (14). To minimize the risk for contamination, we performed template isolation and PCR setup in separate rooms with specified pipette sets, used certified DNAase- and RNase-free filter barrier tips to prevent aerosol contamination, and performed all PCR assays with appropriate controls. After amplification and sequencing, we found that 20 patients were infected with R. japonica.

Figure 1

Figure 1. Phylogenetic analysis of Rickettsia species from febrile patients treated at The People’s Liberation Army 990 Hospital for Rickettsia japonica infection, Xinyang, China, March 2014–June 2017 (bold), and reference species. Tree was...

The gltA gene sequence (456-bp) obtained from all 20 samples (GenBank accession no. MF693145) showed 100% similarity to the corresponding gene of R. japonica strain YH (GenBank accession no. AP017602). The htrA gene sequence (394-bp) obtained from 18 samples (GenBank accession no. MF693146) was identical to that of R. japonica YH, and the sequence in the remaining 2 samples (GenBank accession no. MF693147) differed by only 3 bp. The ompA gene sequence (311-bp) found in 16 samples (GenBank accession no. MF693148) was identical to that of R. japonica YH, and the remaining 4 samples each differed by just 1 bp (GenBank accession nos. MF693149–52) (Figure 1).

We also assessed other tickborne agents, including Anaplasma phygocytophilum, Anaplasma capra, Ehrlichia chaffeensis, Borrelia burgdorferi sensu lato, Babesia microti, and severe fever with thrombocytopenia syndrome virus, by PCR or reverse transcription PCR. Six of 20 patients were co-infected with severe fever with thrombocytopenia syndrome virus, so we excluded them from clinical analysis.

We tested the acute serum samples collected <7 days after illness onset and the convalescence serum samples collected >14 days after illness onset from the 14 patients with R. japonica monoinfection in parallel using an indirect immunofluorescence assay for Rickettsia heilongjiangensis IgG. We considered an IgG titer >1:64 a positive reaction. Using this definition, we found that 6 patients had a seroconversion, and the other 8 had a 4-fold increased IgG titer; thus, all patients had an acute infection with SFG rickettsiae.

All 14 patients had a history of field activity within the previous month, and 5 had a history of tick bite. Median patient age was 61.5 (range 44–77) years; 9 were men. The median time from tick bite to onset of illness was 4 (range 3–8) days and from onset of illness to physician visit 5 (range 2–7) days. The median duration of hospitalization was 6 (range 4–10) days.

Figure 2

Figure 2. Lesions on patients with Rickettsia japonica infection, Xinyang, China, March 2014–June 2017. A) Rash on ventrum; B) rash and eschar on back; C) eschar on femoribus internus; D) tick bite site...

Fever and asthenia were reported by all 14 patients. Other nonspecific symptoms included myalgia (10/14), lymphadenopathy (4/14), headache (3/14), dizziness (3/14), and chills (2/14). Gastrointestinal manifestations included anorexia (9/14), nausea (7/14), vomiting (3/14), and diarrhea (1/14). Three patients had cough and pneumonia, and we observed rash in 10 patients (Figure 2). The median time from onset of illness to rash was 4 (range 3–5) days, and the median duration of rash was 5.5 (range 4–8) days. Only 3 patients had an eschar. Other signs included splenomegaly (2/14) and facial edema (1/14).

Urinalysis on admission revealed microscopic hematuria in 2 patients and a slight or moderate proteinuria in 8 patients. The most common findings on laboratory tests were thrombocytopenia, elevated hepatic aminotransferase concentrations, elevated serum lactate dehydrogenase, and hypoalbuminemia, followed by hyponatremia, anemia, hyperbilirubinemia, hypopotassemia, leukopenia, and elevated serum creatine kinase (Table). Mild multiple organ dysfunction developed in 3 patients.

Because of the appearance of typical rash, we suspected SFG rickettsioses in 10 patients and treated them with doxycycline (100 mg 2×/d) until their fevers dissipated and clinical disease improved (median 4 [range 3–5] days). The other 4 patients were treated with the antimicrobial drugs cefminox or cefoperazone for a median of 6 (range 5–7) days. By the time patients were discharged from the hospital, their leukocyte and platelet counts had returned to reference range levels, but half of the patients still had laboratory test values outside of their respective reference ranges (Table). No patients reported clinically significant sequelae at their 2-week follow-up appointment.

The tick vectors of R. japonica (i.e., H. flava, H. hystricis, H. cornigera, H. longicornis, I. ovatus) are widely distributed throughout China (15), providing R. japonica ample opportunity to infect humans. By applying molecular screening techniques to simultaneously amplify 3 genes, we identified R. japonica infection in 14 patients.

The clinical signs and symptoms of R. japonica infection in our patient cohort differed from those reported in patients in Japan (3,4). We frequently observed fever, asthenia, and rash but not chills and headache. We also saw fewer eschars in our patient cohort, potentially because they were underreported during clinical examination; eschars are not always easy to identify. In addition, the patients we identified usually had myalgia and gastrointestinal symptoms.

These clinical findings expand the available knowledge of the disease spectrum of R. japonica infection. Compared with endemic rickettsiosis caused by Candidatus Rickettsia tarasevichiae infection (reference 16 in Appendix), rash is commonly seen, and hemorrhagic and neurologic signs and symptoms are rarely seen in patients with R. japonica infection. These distinctive features could be used to make a differential diagnosis.

We frequently observed in our patient cohort thrombocytopenia, hypoalbuminemia, elevated hepatic enzyme activity, and elevated lactate dehydrogenase levels, findings resembling those of patients with common SFG rickettsioses (reference 17 in Appendix). In general, disease is mild or moderate, and no deaths have been recorded, although mild multiple organ dysfunction developed in several patients.

In conclusion, we identified R. japonica as an emerging tickborne pathogen in China. Physicians in areas where H. longicornis and other competent vectors for R. japonica are endemic should be aware of the risk for infection in humans and prescribe doxycycline to patients in cases of ineffective therapy with other antimicrobial drugs. Surveillance needs to be extended to improve our understanding of the health burden of JSF.