Nosocomial Transmission of Human Granulocytic Anaplasmosis in China

Home , Boutonneuse fever, Chlamydia, Ixodes persulcatus, Mycoplasma

ORIGINAL CONTRIBUTION

Lijuan Zhang, MD, PhD Context Human granulocytic anaplasmosis (HGA) is an emerging tick-borne disease in Yan Liu, MD China. A cluster of cases among health care workers and family members following ex- Daxin Ni, MD posure to a patient with fulminant disease consistent with HGA prompted investigation. Objective To investigate the origin and transmission of apparent nosocomial cases Qun Li, MD of febrile illness in the Anhui Province. Yanlin Yu, MD Design, Setting, and Patients After exposure to an index patient whose fatal ill- Xue-jie Yu, MD, PhD ness was characterized by fever and hemorrhage at a primary care hospital and re- Kanglin Wan, MD, PhD gional tertiary care hospital’s isolation ward, secondary cases with febrile illness who were suspected of being exposed were tested for antibodies against Anaplasma phago- Dexin Li, MD cytophilum and by polymerase chain reaction (PCR) and DNA sequencing for A phago- Guodong Liang, MD cytophilum DNA. Potential sources of exposure were investigated. Xiugao Jiang, MD Main Outcome Measure Cases with serological or PCR evidence of HGA were compared with uninfected contacts to define the attack rate, relative risk of illness, Huaiqi Jing, MD and potential risks for exposure during the provision of care to the index patient. Jing Run, MD Results In a regional hospital of Anhui Province, China, between November 9 and Mingchun Luan, MD 17, 2006, a cluster of 9 febrile patients with leukopenia, thrombocytopenia, and el- Xiuping Fu, MD evated serum aminotransferase levels were diagnosed with HGA by PCR for A phagocytophilum DNA in peripheral blood and by seroconversion to A phagocytophilum. Jingshan Zhang No patients had tick bites. All 9 patients had contact with the index patient within 12 Weizhong Yang, MD hours of her death from suspected fatal HGA while she experienced extensive hemorrhage and underwent endotracheal intubation. The attack rate was 32.1% vs 0% Yu Wang, MD, PhD (P=.04) among contacts exposed at 50 cm or closer, 45% vs 0% (P = .001) among J. Stephen Dumler, MD those exposed for more than 2 hours, 75% vs 0% (PϽ.001) among those reporting Zijian Feng, MD contact with blood secretions, and 87.5% vs 0% (P=.004) among those reporting contact with respiratory secretions from the index patient. Jun Ren, MD Conclusion We report the identification of HGA in China and likely nosocomial trans- Jianguo Xu, MD, PhD mission of HGA from direct contact with blood or respiratory secretions. UMAN GRANULOCYTIC ANAPLAS- JAMA. 2008;300(19):2263-2270 www.jama.com mosis (HGA) is an emerging tick-borne infectious disease miologicaldatasuggestthatinfectionrates about HGA is limited, the disease is likely that was recognized in the in endemic areas are as high as 15% to underrecognized and underreported UnitedH States in 19901 and in Europe in 36%,6-8 implying that the diagnosis is of- worldwide.7 1997.2 The disease name was changed ten missed or that infection is mild or Despite the pathogen’s global distri- from human granulocytic ehrlichiosis to asymptomatic. Because epidemiological, bution, only a limited number of labo- HGA in 2001 when the causative rickett- clinical, and microbiological information ratory-confirmed cases have been re- sia was reclassified from the genus ehrlichia as Anaplasma phagocytophilum.3 Al- Author Affiliations: National Institute of Communi- Key Laboratory for Infectious Diseases Prevention and cable Disease Control and Prevention, China CDC, Bei- Control, Beijing, China (Drs Wan, D. Li, Liang, Jing, though the clinical presentation of HGA jing (Drs L. Zhang, Wan, Jiang, Jing, Luan, Fu, and and Xu); and Department of Pathology, The Johns is variable and although it may be diffi- Xu and Mr J. Zhang); Anhui Center for Disease Con- Hopkins University School of Medicine, Baltimore, trol and Prevention, Hefei, Anhui Province, China (Drs Maryland (Dr Dumler). cult to diagnose, the annual number of in- Liu, Q. Li, and Ren); Chinese Center for Diseases Pre- Corresponding Author: Jianguo Xu, MD, PhD, National fectionsreportedintheUnitedStatessince vention and Control, Beijing (Drs Ni, Yang, Wang, and Institute of Communicable Disease Control and Pre- 4,5 Feng); Yijishan Hospital, Wuhu, Anhui Province, China vention, China, CDC, Beijing, China, PO Box 5, Chang- 1990 has steadily increased. Seroepide- (Drs Yu and Run); Department of Pathology, Univer- ping, Beijing, China 102206 ([email protected]) and sity of Texas Medical Branch, Galveston (Dr X. Yu); Jun Ren, MD, Anhui Center for Disease Control and For editorial comment see p 2308. National Institute for Viral Diseases Prevention and Prevention, He Fei, Anhui Province, 230061, China Control, Beijing, China CDC (Drs D. Li and Liang); State ([email protected]).

Downloaded From: https://jamanetwork.com/ on 09/26/2021 HUMAN GRANULOCYTIC ANAPLASMOSIS IN CHINA

ported from countries in Europe, where rus, yellow fever virus, Crimean- perience with tick bites; exposure to the the median seroprevalence rate is 6.2%, Congo hemorrhagic fever virus, cox- index patient—where, when, and how similar to that in North America.9 Se- sackievirus, respiratory syncytial virus, they had contact; exposure to wild ani- rological and molecular evidence also adenovirus, Mycoplasma pneumoniae, mals; extent of outdoor activity; expo- suggests that human infection exists in Chlamydia species, Ehrlichia species, sure to the index patient’s blood and Korea, Japan, and China.10-14 Herein, we Rickettsia species, and Orientia respiratory secretions or to grossly report the first cases of HGA acquired tsutsugamushi. bloody oropharyngeal secretions; pres- in China, as well as the unusual find- Tests were also conducted on oro- ence of skin lesions during exposure; ing of nosocomial human-to-human pharygeal swabs from the first 3 to 5 whether skin surfaces were washed af- transmission. days after onset for influenza A virus ter exposure; whether they were ex- antigens, and by PCR for influenza A posed to the patient’s stool or urine; and METHODS viruses, influenza B virus, and influ- the timing of these events. Health care Laboratory Diagnosis enza virus subtype H5 nucleic acids. workers were asked about their use of Patients suspected of HGA exposure Tests for acute-phase serum were con- masks and gloves. were tested for serum IgG to A phago- ducted to detect IgM and IgG to se- cytophilum using the IgG IFA kit (Fo- vere acute respiratory syndrome vi- Ethical and Human Subjects cus Diagnostics, Cypress, California), rus, as well as to detect IgM or IgM plus Review screening at a 1:64 dilution and titrat- IgG antibodies by capture enzyme- The study was approved by the ethics ing if reactive.15 Nested polymerase linked immunosorbent assay against committee of China CDC, according to chain reaction (PCR) using blood DNA Bunyaviridae, Filoviridae, Lassa fever the medical research regulations of Min- (QIAamp DNA Mini Kit, QIAGEN, virus, Ebola virus, Marburg virus, Han- istry of Health, China. Oral informed Hilden, Germany) was used to detect taan virus, Junin virus, yellow fever vi- consent was obtained from all study A phagocytophilum DNA with Ana- rus, and Crimean-Congo hemor- participants. plasma and Ehrlichia genus-common rhagic fever virus. and A phagocytophilum species- Statistical Analysis specific rrs primers (16S rRNA gene),16 Epidemiological Investigation All statistical calculations were per- and A phagocytophilum species- All contacts of the index patient, informed using Epi Info 6.04d (http: specific groEL primers.17 An A phago- cluding patients with similar clinical //www.cdc.gov/epiinfo). To identify cytophilum rrs plasmid and DNA from presentations and healthy persons, were specific exposure risk factors, retro- healthy people or distilled water were interviewed before laboratory diagnos- spective cohort comparisons were used as controls. Positive reactions were tic results were obtained. A possible case evaluated by calculating attack rates, confirmed by direct sequencing. Poly- of HGA was defined as a patient with a relative risk, and 95% confidence in- merase chain reaction was conducted clinically compatible illness (fever, tervals and by Fisher exact test; signifi- in 2 independent laboratories, the Na- headache, chills) and laboratory find- cance was defined as a 2-tailed P Ͻ.05. tional Institute for Communicable Dis- ings including thrombocytopenia and ease Control and Prevention in Bei- leukopenia but who lacked serologi- RESULTS jing, and at the Anhui Province Center cal or molecular tests for A phagocyto- Index Case for Disease Prevention and Control in philum. A confirmed case was defined A 50-year-old woman with a 1-day Hefei city. Each laboratory used its own as a patient with a clinically compat- abrupt onset of sudden fever (39.2°C), primers, reagents, and patient blood ible illness (as above) and in keeping headache, myalgia, arthralgia, dizzi- DNA. All samples were tested concur- with the US Centers for Disease Con- ness, and malaise presented to the vil- rently with negative and no template trol and Prevention (CDC) criteria lage clinic on October 31, 2006, and was controls (water) under the same con- (http://www.cdc.gov/ncphi/disss/nndss treated with ribavirin, cephalothin, dexa- ditions. Polymerase chain reaction /casedef/ehrlichiosis_2008.htm) by methasone, and amidopyrine for 4 days. samples from healthy people and nega- either seroconversion, a 4-fold in- At 9 PM on November 3, she was admit- tive controls consistently had nega- crease in A phagocytophilum IgG anti- ted to the local hospital because of gum tive results. body titer in acute and convalescent bleeding, facial edema, nausea, vomit- To exclude other infections, sero- sera, or a positive PCR result for both ing, and oliguria, a temperature of logical, antigen detection, and PCR di- A phagocytophilum rrs and groEL con- 39.7°C, blood pressure of 85/60 mm Hg, agnostic tests were conducted. These firmed by direct sequence analysis.15 and pulse rate of 96/min; a rash was included tests on blood from the first noted over her trunk. Laboratory test- 3 to 5 days after onset for reverse tran- Contact Questionnaire ing showed leukopenia (white blood cell scription (RT)–PCR of PCR for nucleic All contacts of the index patient were count, 3300/µL), thrombocytopenia acids of Lassa fever virus, Ebola virus, asked to complete a questionnaire about (platelet count, 18ϫ103/µL), elevated Marburg virus, Hantaan virus, Junin vi- their health status and profession; ex- serum aspartate aminotransferase

Downloaded From: https://jamanetwork.com/ on 09/26/2021 HUMAN GRANULOCYTIC ANAPLASMOSIS IN CHINA

(629 U/L) and alanine aminotransfer- (9 of 9 patients), myalgia (5 of 9), in the final 12 hours of her life while she ase (69 U/L), elevated creatinine (2.6 mg/ diarrhea (7 of 9), leukopenia (white was in the critical care unit and during dL), and elevated blood urea nitrogen blood cell count, 1200-3700/µL in 9 the endotracheal intubation proce- (48 mg/dL) levels. Dipstick urinalysis re- of 9), thrombocytopenia (platelets, dure. No one whose only contact with vealed 3ϩ hematuria and 3ϩ protein- 39-115ϫ103/µL in all 9), and elevated the index patient was before these 12 uria (protein, 3 g/L). (To convert aspar- serum aspartate aminotransferase and hours was infected. tate aminotransferase and alanine alanine aminotransferase (7 of 9) aminotransferase to microkat per liter, (TABLE 1). All patients had contact Serological and Molecular multiply by 0.0167; creatinine to mi- with the index patient, including 5 Diagnosis cromole per liter, by 88.4; and urea ni- family members, 2 physicians, and 2 Anaplasma phagocytophilum IgG sero- trogen to millimole per liter, by 0.357.) nurses who had accompanied or conversions were detected for all 9 pa- Her condition progressively deterio- treated her between November 4 and tients, and a 4-fold IgG titer increase rated, so she was transferred to a re- 5 (Figure). was observed in 7 of 9 patients gional hospital at 11 AM, November 4. The initial secondary case experi- (Table 1). Nested PCR using genus- By 7 PM, the patient became ob- enced fever on November 9, 4 days af- common rrs and species-specific rrs and tunded, cyanotic, and purpuric and was ter death of the index case, followed on groEL primers identified A phagocyto- bleeding from her nose and mouth. This November 11 by another patient, on philum DNA in the blood samples from extensive mouth and nose bleeding re- November 12 by 3 patients, and on No- all 9 patients when they were in the quired frequent aspiration and con- vember 14 by 3 more patients. The last acute phase, whereas all healthy and taminated the working area surfaces, patient reported illness on November template controls had negative test re- health care workers, and family mem- 17, 12 days after the death of the in- sults. The identity of amplicons from bers who were with her. Family mem- dex patient. The patients were be- each of the 9 patients was confirmed by bers assisted with patient care by wip- tween 25 and 67 years (mean, 36.2 sequencing; all rrs sequences (206 base ing blood from the patient’s mouth and years), and 6 were men. All were pre- pairs) were identical and all groEL se- nose, rinsing and reusing the same tow- viously healthy. The average incuba- quences (446 nt) were identical (Gen- els. By 7:38 PM the patient developed tion period was 7.8 days (range, 4-12 Bank accession numbers: rrs rapidly progressive dyspnea and wors- days). All had fever of at least 38.5°C EF211110-17 and EF473210; groEL ening oxygen saturation and required for 1 to 6 days (mean, 4 days). Diar- EF47320108 and EF473209). Al- endotracheal intubation. The patient re- rhea was characterized as 1 to 3 loose though the rrs sequences were identi- mained hypoxic and hypotensive with stools per day persisting for 1 to 2 days. cal to most other human-derived strains multiorgan failure and copious bleed- All patients had relative bradycardia. globally, sequences from groEL were ing from the nose and mouth. Despite One patient developed acute respira- identical to some US strains (Wiscon- all efforts, the patient died at 6:45 AM, tory distress syndrome as a complica- sin and New York) but differed from A November 5, 2006. The final diagno- tion of Aspergillus pneumonia and tu- phagocytophilum in China (93.6%; sis was hemorrhagic fever with renal berculosis during his hospitalization. EU008083), Germany (99.4%; syndrome, even though no IgG anti- The other 8 patients were mildly af- AY281850), and California (99.7%; bodies to Hantaan virus were de- fected, recovered, and were dis- U96727). These data support the tected. A postmortem examination was charged in good health. premise that a single clone was respon- not performed, and no blood or tissue sible for all of the 9 secondary cases. Al- samples remained for retrospective Contact Investigation though peripheral blood smears were laboratory testing. The index patient had contact with 63 examined for all 9 patients at the time Retrospective questioning of the pa- persons after onset of illness: 21 family of illness, no convincing evidence of A tient’s family revealed that she was bit- members and 42 health care workers. Of phagocytophilum morulae was ob- ten by a tick 12 days before onset of fe- the 42 health care workers, 18 were from served. All RT-PCR, PCR, antigen de- ver: she had killed several mice in her the local hospital, including 2 from the tection, and IgM antibody detection home 9 days before onset, and her hus- village clinic, and 24 were from the re- tests for microbial and viral etiologies band had hunted and brought home gional hospital. Of the 21 family mem- were negative. “wild animal carcasses” 3 days before bers, 4 had contact with the index pa- onset of illness. A timeline of events is tient in only the local hospital, 13 only Risk Factors shown in the FIGURE. in the regional hospital, and 4 in both. The exposure data implicate transmis- The 9 secondary cases occurred among sion at the regional hospital, permit- Nosocomial Cases of HGA the 39 health care workers and relating focus on risk factors in 39 indi- Between November 9 and 17, 2006, 9 tives with patient exposure at the re- viduals, including 24 health care patients were identified at the regional gional hospital, representing an attack workers and 15 family members hospital with fever higher than 38.0°C rate of 23%. All 9 cared for the index case (TABLE 2). Two family members who

Downloaded From: https://jamanetwork.com/ on 09/26/2021 HUMAN GRANULOCYTIC ANAPLASMOSIS IN CHINA

Figure. Timeline of Critical Events for the Index Patient and Direct Contact Intervals of Family Members and Health Care Workers With the Index Patient and Exposure of Patients With Nosocomial Human Granulytic Anaplasmosis

Index patient events Index patient Transfer to regional hospital Family member Local hospital Intubation, hemorrhage, Health care worker admission life support Killed mice Exposed to wild Fever Village 3 Tick bite in home animal carcasses onset clinic visit Death E H 2 D G 1 ABCFJ 0 No. of Cases 1617 18 192123242526272829303112345678910111213141516 20 22 17 18 October November

Hospitalized in Transferred to Hemorrhage local hospital regional hospital Intubation Hemorrhage Death

6 PM 12 AM 6 AM 12 PM 6 PM 12 AM 6 AM 12 PM Exposed November 3 November 4 November 5 individuals Family members 4 Individuals (uncertain exposure period; not infected) (840) (730) (420) (330) (5) (15) (40) (40) (30) (120) (5) (5) F (730) C (850) B (175) D (120) A (190) Local health 18 Individuals care workers (uncertain exposure period; not infected)

Regional health (5) care workers (5) (5) (600) (5) (5) (30) (240) (300) (240) (240) (180) (8) (30) Potential exposure period (30) (3) Time uncertain (not infected) (5) (2) Not infected (5) (10) Infected G (270) Family member E (180) Health care worker H (200) J (180)

6 PM 12 AM 6 AM 12 PM 6 PM 12 AM 6 AM 12 PM Period of potential exposure (total minutes of exposure)

Top, epidemic curve showing progression of outbreak and key events during the index patient’s illness. Bottom, each bar indicates the period of potential exposure while family members were in the hospital and while health care workers were assigned to care for the index patient. Duration of exposure in minutes is shown in parentheses and may not have occurred continuously during the exposure period. Capital letters designate the corresponding secondary cases in the top and bottom panels.

Downloaded From: https://jamanetwork.com/ on 09/26/2021 HUMAN GRANULOCYTIC ANAPLASMOSIS IN CHINA

had contact with the index case after from the regional hospital wore masks patient during the same time was her death were not included. and 9 of 24 (38%) wore gloves. infected. The index patient was None of the 9 secondary cases re- Of 17 family members who exposed to 20 contacts for more than ported tick bites, exposure to wild ani- reported contact with the index 2 hours, and 9 were infected, whereas mals, or participation in hunting activ- patient at the regional hospital, 13 none of 19 contacts exposed fewer ity in the preceding 2 months, and only were present during endotracheal than 2 hours was infected. All 9 1 reported recent outdoor activity. For intubation, 5 of whom were infected. infected patients reported contact all 9 secondary cases, culture serologi- Of these 5 individuals, 3 reported with blood (P = .002) and 7 had con- cal, antigen detection, and nucleic acid blood contamination of skin and pos- tact with respiratory secretions (rela- detection studies for other infectious sible mucocutaneous exposures, sug- tive risk, 7.0; 95% confidence interval, etiologies were negative. gesting direct contact with blood or 1.7-29.1; Table 2). Those persons Of 24 regional hospital health care respiratory secretions as the mecha- with skin expsoure to blood workers who had contact with the in- nism of transmission. (PϽ.001) or respiratory secretions dex patient, 18 were on duty during the Among the 28 individuals who (P = .004), or those with preexisting final 12 hours, and 4 of the 18 who were reported close contact (Յ50 cm) with skin lesions or injuries followed by involved in the endotracheal intuba- the index patient during the final 12 exposure to blood (relative risk, 3.6; tion were infected. Of these 4, 3 were hours of her life, 9 were infected. In 95% confidence interval, 1.1-7.6; involved in endotracheal intubation and contrast, none of the 11 individuals P=.02) were significantly more likely care during times of hemorrhage. Six- who reported a physical distance of to be infected (TABLE 3). Neither teen of 24 health care workers (67%) more than 50 cm from the index exposure to stool nor exposure to

Table 1. Clinical, Laboratory, and Serological Findings of 9 Patients With Nosocomial Human Granulocytic Anaplasmosis Infected Patients

2345678910 Clinical findingsa Days hospitalized 19 21 19 19 19 19 21 19 36 Temperature Ն38.5°C Yes Yes Yes Yes Yes Yes Yes Yes Yes Malaise Yes Yes Yes Yes Yes Yes Yes Yes Yes Chills Yes Yes Yes Yes Yes Yes Yes Yes Yes Diarrhea Yes No Yes Yes No Yes Yes Yes Yes Myalgia Yes No Yes Yes No No Yes No Yes Coryza/pharyngitis No No No No Yes Yes No No Yes Headache Yes No No No No Yes No No No Nausea No No Yes No No No No No Yes Edema No No No No No Yes No No No Gum bleeding No No No No No Yes No No No Dysuria No No No No No No No No Yes Laboratory values Lowest blood count, range of normal White blood cell, 2600 1900 2700 2100 2500 1200 1800 3700 2200 4500-11 000/µLa Platelet, 46 49 85 39 115 47 40 52 42 150-350 ϫ 103/µL Highest liver enzymes, U/L AST, men Ͻ38; women Ͻ32 252 116 ND 77 ND 50 50 77 78 ALT, men Ͻ40; women Ͻ31 84 66 ND 64 ND 89 89 74 139 Anaplasma phagocytophilum IgG titers Days after onset 0-7 Ͻ64 Ͻ64 Ͻ64 64 64 Ͻ64 Ͻ64 Ͻ64 Ͻ64 20-25 ND 64 64 128 128 128 ND 64 128 55-70 256 256 Ͻ64 256 256 Ͻ64 64 128 ND A Phagocytophilum PCR results rrs Yes Yes Yes Yes Yes Yes Yes Yes Yes groEL Yes Yes Yes Yes Yes Yes Yes Yes Yes Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; ND, not done; PCR, polymerase chain reaction. a Clinical findings that were documented during the course of each patient’s hospitalization.

Downloaded From: https://jamanetwork.com/ on 09/26/2021 HUMAN GRANULOCYTIC ANAPLASMOSIS IN CHINA

urine from the index case resulted in with hemorrhage. Although the index ries.18,19 Infection can be severe, with increased risk (0.6 and 1.1, respec- patient can only be categorized as a intensive care unit admission required tively). possible case, clinical and historical in 7% of patients and fatalities occur- support for the diagnosis of HGA is ring in up to 1%, yet most infections COMMENT strong. She had a tick bite within are sporadic and probably self- Nine cases of A phagocytophilum the known incubation period and limited.4 Based on the mild to moder- infection were confirmed at the had a clinical presentation compatible ate severity observed in 8 of the 9 sec- regional hospital in the Anhui Prov- with severe HGA.4 Moreover, the ondarily infected patients, Chinese ince of China in a 9-day period. All epidemiological investigation of HGA conforms to the spectrum of presented with HGA as described in exposed individuals with HGA impli- clinical severity observed in North North America and Europe7 and ful- cates her as the index case. Unfortu- America.4,7,15 The fatal outcome in the filled the US CDC laboratory criteria nately, no tissue or serum sample is index case is clinically similar to that for the diagnosis of HGA.15 The most available to confirm retrospectively observed for other HGA fatalities, remarkable aspect of these cases was her diagnosis. including exsanguination with sepsis that transmission was very unlikely to Human granulocytic anaplasmosis syndrome possibly relating to cyto- be tick-borne, but was closely associ- and human monocytic ehrlichiosis kine overproduction, opportunistic ated with blood or respiratory secre- were initially identified with presenta- infections, and increased HGA sever- tion exposure from an index patient tions now recognized as relatively ity in the setting of preexisting medi- who died of a fulminant febrile illness uncommon for their natural histo- cal conditions such as diabetes melli- tus.7,20 A phagocytophilum transmission in Table 2. Risk Factors for Acquisition of Human Granulocytic Anaplasmosis Among 39 Contacts Exposed to Index Patient While at the Regional Hospital China and Asia is predicated on the No./Total (%) presence of this zoonotic agent in vector ticks and vertebrate hosts. Attack Rate Attack Rate Relative Risk Although studies in Asia are limited, With Exposure Without Exposure (95% Confidence P Exposure to Index Patient Factor Factor Interval)a Valueb at least 8 have examined A phagocyto- Յ50 cm to nose and mouth 9/28 (32.1) 0/11 (0) .04 philum infection of ticks, including Ͼ2 h 9/20 (45.0) 0/19 (0) .001 2284 Ixodes persulcatus ticks, of which During or after intubation 9/30 (30.0) 0/9 (0) .09 4.4% carried A phagocytophilum DNA, During massive hemorrhage 4/9 (44.4) 5/30 (16.7) 2.7 (0.9-7.9) .17 a prevalence similar to that in Euro- period pean and North American Ixodes spe- Any direct blood contact 9/22 (40.9) 0/17 (0) .002 cies ticks.12,14,21-27 Likewise, 9% and Direct respiratory or tracheal 7/13 (53.8) 2/26 (7.7) 7.0 (1.7-29.1) .003 24% of Apodemus species field mice in secretion contact aInfinite or not able to be calculated. northern China and Korea, respec- bFisher exact test (2 tailed). tively, and 64% of Crosidura lasiura shrews in Korea are infected.13,21,24,28,29 Although no proven cases of HGA Table 3. Risk Factors for Human Granulocytic Anaplasmosis Associated With Direct Exposure have been previously identified in to Index Patient’s Blood and Respiratory Secretions China, at least 1 study describes A No./Total (%) phagocytophilum DNA in the blood of 4 Chinese patients with tick bites,14,30 Attack Rate Attack Rate Relative Risk With Exposure Without Exposure (95% Confidence P and seroepidemiological investiga- Exposure Factor Factor Factor Interval)a Valueb tions demonstrate that 2% to 9% of Any direct blood contact febrile patients in Korea,10,11 and during hemorrhage On skin 9/12 (75.0) 0/10 (0) Ͻ.001 between 0.5% and 6% of healthy Chi- Open wounds 4/4 (100.0) 5/18 (27.8) 3.6 (1.1-7.6) .02 nese residents have A phagocytophilum or abrasions antibodies.31 Not washed timely 4/8 (50.0) 5/14 (35.7) 1.4 (0.5-3.8) .66 Rare examples of nontick transmis- Direct respiratory or tracheal sion of HGA exist in the literature secretion contact On skin 7/8 (87.5) 0/5 (0) .004 and include direct exposure to deer 32 33 Open wounds 4/4 (100.0) 3/9 (33.3) 3.0 (1.2-7.6) .07 blood, transfusion, and transpla- or abrasions cental transmission.34 Similarly, under Not washed timely 3/6 (50.0) 4/7 (57.1) 0.9 (0.3-2.4) Ͼ.99 the proper circumstances other rick- aInfinite or not able to be calculated. ettsial infections are transmissible via bFisher exact test (2-tailed). aerosol, direct contact with mucous

Downloaded From: https://jamanetwork.com/ on 09/26/2021 HUMAN GRANULOCYTIC ANAPLASMOSIS IN CHINA

membranes or conjunctivae, or pital and health care unit operations to MMWR Morb Mortal Wkly Rep. 2007;54(53): 35-38 2-92. mechanical fomite transmission. prevent any additional nosocomial out- 6. Bakken JS, Goellner P, Van Etten M, et al. Sero- Direct exposure to small blood vol- breaks of HGA. Moreover, as China ad- prevalence of human granulocytic ehrlichiosis among permanent residents of northwestern Wisconsin. Clin umes probably carries a low risk vances into its future, it must also now Infect Dis. 1998;27(6):1491-1496. because experimental and natural become prepared to deal with the in- 7. Dumler JS, Choi KS, Garcia-Garcia JC, et al. Hu- infections of white-tailed deer result creasing threat that tick-borne rickett- man granulocytic anaplasmosis and Anaplasma 39 phagocytophilum. Emerg Infect Dis. 2005;11(12): in only low-level bacteremia. How- sial pathogens have been already 1828-1834. ever, it is possible that this low risk brought to the United States and Eu- 8. IJdo JW, Meek JI, Cartter ML, et al. The emer- gence of another tickborne infection in the 12-town may be offset by large volumes of ani- rope. area around Lyme, Connecticut: human granulocytic mal blood and tissues, such as those Author Contributions: Dr Xu had full access to all of ehrlichiosis. J Infect Dis. 2000;181(4):1388-1393. to which butchers are exposed. the data in the study and takes responsibility for the 9. Dumler JS. Anaplasma and ehrlichia infection. Ann integrity of the data and the accuracy of the data analy- N Y Acad Sci. 2005;1063:361-373. Another factor related to transmis- sis. Drs Zhang, Liu, Ni, Li, Y. Yu, and X. Yu contrib- 10. Heo EJ, Park JH, Koo JR, et al. Serologic and mo- sibility is the blood burden of A uted equally to this work. lecular detection of Ehrlichia chaffeensis and Ana- Study concept and design: L. Zhang, Liu, Ni, Li, Y. Yu, plasma phagocytophila (human granulocytic ehrli- phagocytophilum, which appears to Wan, Jing, Rui, Yang, Wang, Dumler, Feng, Ren, Xu. chiosis agent) in Korean patients. J Clin Microbiol. increase with immunosuppression Acquisition of data: Liu, Ni, D. Li, Y. Yu, Wan, Q. Li, 2002;40(8):3082-3085. Liang, Jiang, Jing, Rui, Luan, Fu, J. Zhang, Xu. 11. Park JH, Heo EJ, Choi KS, et al. Detection of an- resulting in absolute infected neutro- tibodies to Anaplasma phagocytophilum and Ehrli- 9 Analysis and interpretation of data: L. Zhang, Liu, Ni, phil counts as high as 2.7 to 5.9ϫ10 / Li, Y. Yu, X. Yu, Wan, Liang, Jiang, Jing, Dumler, Feng, chia chaffeensis antigens in sera of Korean patients L.18,40 It is unclear to what degree the Xu. by western immunoblotting and indirect immunofluo- Drafting of the manuscript: L. Zhang, Liu, Ni, Q. Li, rescence assays. Clin Diagn Lab Immunol. 2003; sustained dexamethasone treatment of Y. Yu, Wan, Liang, Jiang, Jing, Luan, Fu, J. Zhang, 10(6):1059-1064. the index case contributed to trans- Dumler, Xu. 12. Ohashi N, Inayoshi M, Kitamura K, et al. Ana- Critical revision of the manuscript for important in- plasma phagocytophilum–infected ticks, Japan. Emerg mission. The final consideration is the tellectual content: X. Yu, Q. Li, Rui, Yang, Wang, Dum- Infect Dis. 2005;11(11):1780-1783. likelihood of health care worker and ler, Feng, Ren, Xu. 13. Chae JS, Kim CM, Kim EH, et al. Molecular epi- family member exposure to sufficient Statistical analysis: L. Zhang, Ni, Q. Li, Y. Yu, Wan, demiological study for tick-borne disease (Ehrlichia and Liang, Jiang, Jing, Rui, Luan, Fu, Zhang. Anaplasma spp.) surveillance at selected US military volumes of infectious body fluids to Obtained funding: Rui, Wang, Xu. training sites/installations in Korea. Ann N Y Acad Sci. account for transmission. It is not Administrative, technical, or material support: Liu, Ni, 2003;990:118-125. D. Li, Y. Yu, X. Yu, Wan, D. Li, Liang, Jiang, Jing, Rui, 14. Wen B, Cao W, Pan H. Ehrlichiae and ehrlichial unusual for occupational blood expo- Yang, Feng, Ren, Xu. diseases in China. Ann N Y Acad Sci. 2003;990: sure to occur among those caring for Study supervision: Wang, Dumler, Feng, Xu. 45-53. Financial Disclosures: Dr Dumler reports that he holds 15. Chapman AS, Bakken JS, Folk SM, et al. Diagno- patients with hemorrhage or during a patent for a method for in vitro propagation of A sis and management of tickborne rickettsial diseases: procedures such as intubation or sur- phagocytophilum for which royalty fees are paid. Oth- Rocky Mountain spotted fever, ehrlichioses, and ana- gery, for which the relative risk is 3 to erwise no other authors report disclosures of finan- plasmosis—United States: a practical guide for phy- cial or potential conflicts of interest. sicians and other health-care and public health 4 times higher than for other medical Funding/Support: This work was supported by grants professionals. MMWR Recomm Rep. 2006;55 specialties.41 In western societies, 2005CB522904 and 200802016 (Dr Xu) from the Min- (RR-4):1-27. istry of Science and Technology and by emerging re- 16. Wen B, Jian R, Zhang Y, et al. Simultaneous de- most family members are excluded sponse found from National Institute of Communi- tection of Anaplasma marginale and a new Ehrlichia from these events and health care cable Disease Control and Prevention, China CDC, species closely related to Ehrlichia chaffeensis by se- Beijing, China. quence analyses of 16S ribosomal DNA in Boophilus workers are increasingly protected by Role of the Sponsor: The sponsors provided funding, microplus ticks from Tibet. J Clin Microbiol. 2002; training and barriers such as gloves, but had no role in determining study design, data col- 40(9):3286-3290. gowns, and masks.42 However, retro- letion, or interpretation. 17. Bell CA, Patel R. A real-time combined polymerase chain reaction assay for the rapid detection and spective questioning of our cases differentiation of Anaplasma phagocytophilum, Ehrli- REFERENCES clearly indicated that both family chia chaffeensis,andEhrlichia ewingii. Diagn Micro- members and health care workers not 1. Chen SM, Dumler JS, Bakken JS, et al. Identifica- biol Infect Dis. 2005;53(4):301-306. tion of a granulocytotropic Ehrlichia species as the etio- 18. Bakken JS, Dumler JS, Chen SM, et al. Human only participated in these events but logic agent of human disease. J Clin Microbiol. 1994; granulocytic ehrlichiosis in the upper Midwest United were unlikely to use gloves and so 32(3):589-595. States: a new species emerging? JAMA. 1994; 2. Petrovec M, Lotric Furlan S, Zupanc TA, et al. Hu- 272(3):212-218. reported that body surfaces were con- man disease in Europe caused by a granulocytic Ehrli- 19. Maeda K, Markowitz N, Hawley RC, et al. Hu- taminated by potentially infectious chia species. J Clin Microbiol. 1997;35(6):1556- man infection with Ehrlichia canis, a leukocytic 1559. rickettsia. N Engl J Med. 1987;316(14):853-856. fluids. Moreover, many participants 3. Dumler JS, Barbet AF, Bekker CP, et al. Reorgani- 20. Dumler JS, Barat NC, Barat CE, et al. Human granu- did not acknowledge use of postexpo- zation of genera in the families Rickettsiaceae and Ana- locytic anaplasmosis and macrophage activation. Clin plasmataceae in the order Rickettsiales: unification of Infect Dis. 2007;45(2):199-204. sure precautions, such as hand and some species of Ehrlichia with Anaplasma, Cowdria 21. Cao WC, Zhan L, He J, et al. Natural Anaplasma skin washing. with Ehrlichia and Ehrlichia with Neorickettsia, de- phagocytophilum infection of ticks and rodents from Although it is likely that routine scriptions of six new species combinations and des- a forest area of Jilin Province, China. AmJTropMed ignation of Ehrlichia equi and “HGE agent” as sub- Hyg. 2006;75(4):664-668. blood and body fluid precautions will jective synonyms of Ehrlichia phagocytophila. Int J Syst 22. Cao WC, Zhao QM, Zhang PH, et al. Granulo- protect against such future events, strict Evol Microbiol. 2001;51(pt 6):2145-2165. cytic Ehrlichiae in Ixodes persulcatus ticks from an area 4. Dumler JS, Madigan JE, Pusterla N, et al. Ehrlichio- in China where Lyme disease is endemic. J Clin adherence to protective protocols is ses in humans: epidemiology, clinical presentation, di- Microbiol. 2000;38(11):4208-4210. mandatory even if communicability is agnosis, and treatment. Clin Infect Dis. 2007;45 23. Cao WC, Zhao QM, Zhang PH, et al. Prevalence (suppl 1):S45-S51. of Anaplasma phagocytophila and Borrelia burgdor- deemed unlikely. The lessons of this 5. McNabb SJ, Jajosky RA, Hall-Baker PA, et al. Sum- feri in Ixodes persulcatus ticks from northeastern China. study remain relevant to the daily hos- mary of notifiable diseases—United States, 2005. Am J Trop Med Hyg. 2003;68(5):547-550.

Downloaded From: https://jamanetwork.com/ on 09/26/2021 HUMAN GRANULOCYTIC ANAPLASMOSIS IN CHINA

24. Kim CM, Yi YH, Yu DH, et al. Tick-borne rickett- man ehrlichia infectious people in Daxingan Moun- conjunctival, gastric, and subcutaneous routes and pro- sial pathogens in ticks and small mammals in Korea. tains [in Chinese]. Zhonghua Liu Xing Bing Xue Za Zhi. tection afforded by an experimental vaccine. Infect Appl Environ Microbiol. 2006;72(9):5766-5776. 2001;22(2):137-141. Immun. 1979;25(2):580-582. 25. Rar VA, Fomenko NV, Dobrotvorsky AK, et al. Tick- 31. Zhang L, Shan A, Mathew B, et al. Rickettsial se- 38. Hawkins JA, Love JN, Hidalgo RJ. Mechanical trans- borne pathogen detection, Western Siberia, Russia. roepidemiology among farm workers, Tianjin, Peo- mission of anaplasmosis by tabanids (Diptera: Emerg Infect Dis. 2005;11(11):1708-1715. ple’s Republic of China. Emerg Infect Dis. 2008; Tabanidae). Am J Vet Res. 1982;43(4):732-734. 26. Shpynov S, Fournier PE, Rudakov N, et al. De- 14(16):938-940. 39. Tate CM, Mead DG, Luttrell MP, et al. Experi- tection of members of the genera Rickettsia, Ana- 32. Bakken JS, Krueth JK, Lund T, et al. Exposure to mental infection of white-tailed deer with Ana- plasma, and Ehrlichia in ticks collected in the Asiatic deer blood may be a cause of human granulocytic plasma phagocytophilum, etiologic agent of human part of Russia. Ann N Y Acad Sci. 2006;1078:378- ehrlichiosis. Clin Infect Dis. 1996;23(1):198. granulocytic anaplasmosis. J Clin Microbiol. 2005; 383. 33. Leiby DA, Gill JE. Transfusion-transmitted tick- 43(8):3595-3601. 27. Kim CM, Kim MS, Park MS, et al. Identification borne infections: a cornucopia of threats. Transfus Med 40. Bayard-Mc Neeley M, Bansal A, Chowdhury I, et al. of Ehrlichia chaffeensis, Anaplasma phagocytophi- Rev. 2004;18(4):293-306. In vivo and in vitro studies on Anaplasma phagocy- lum, and A bovis in Haemaphysalis longicornis and 34. Horowitz HW, Kilchevsky E, Haber S, et al. Peri- tophilum infection of the myeloid cells of a patient with Ixodes persulcatus ticks from Korea. Vector Borne Zoo- natal transmission of the agent of human granulo- chronic myelogenous leukaemia and human granu- notic Dis. 2003;3(1):17-26. cytic ehrlichiosis. N Engl J Med. 1998;339(6):375- locytic ehrlichiosis. J Clin Pathol. 2004;57(5):499- 28. Zhao QM, Wu XM, Zhang PH, et al. Study on 378. 503. the coinfection of three tick-borne infectious dis- 35. Oster CN, Burke DS, Kenyon RH, et al. Laboratory- 41. Denis MA, Ecochard R, Bernadet A, et al. Risk of eases in China using polymerase chain reaction method acquired Rocky Mountain spotted fever. The hazard occupational blood exposure in a cohort of 24,000 hos- [in Chinese]. Zhonghua Liu Xing Bing Xue Za Zhi. 2005; of aerosol transmission. NEnglJMed. 1977;297 pital healthcare workers: position and environment 26(1):9-13. (16):859-863. analysis over three years. J Occup Environ Med. 2003; 29. Chahan B, Jian Z, Xuan X, et al. Serological evi- 36. Diez Ruı´z A, Ramos Jime´ nez A, Lo´ pez Ruz MA, 45(3):283-288. dence of infection of Anaplasma and Ehrlichia in do- et al. Boutonneuse fever transmitted by conjunctival 42. Tarantola A, Golliot F, Astagneau P, et al. Occu- mestic animals in Xinjiang Uygur Autonomous Re- inoculation. Klin Wochenschr. 1988;66(24):1212- pational blood and body fluids exposures in health care gion area, China. Vet Parasitol. 2005;134(3-4): 1213. workers: four-year surveillance from the Northern 273-278. 37. Kenyon RH, Kishimoto RA, Hall WC. Exposure of France network. Am J Infect Control. 2003;31 30. Gao D, Cao W, Zhang X. Investigations on Hu- guinea pigs to Rickettsia rickettsii by aerosol, nasal, (6):357-363.

Nothing is more estimable than a physician who, hav- ing studied nature from his youth, knows the prop- erties of the human body, the diseases which assail it, the means which will benefit it, exercises his art with caution, and pays equal attention to the rich and the poor. —Voltaire (1694-1778)

Downloaded From: https://jamanetwork.com/ on 09/26/2021