Scrub Typhus Involving Central Nervous System, India, 2004–2006

To the Editor: Scrub typhus, caused by Orientia tsutsugamushi, is one of the most common infectious diseases of rural southern Asia, southeastern Asia, and the western Pacific. The disease is transmitted to humans by the bite of larvae of trombiculid mites harboring the pathogen. The disease often appears as a nonspecific febrile illness. The clinical picture of scrub typhus is typically associated with fever, rash, myalgia, and diffuse lymphadenopathy (1). Immunofluorescence assay (IFA) is the test of choice for serodiagnosis of rickettsial diseases (2). Scrub typhus has been reported from northern, eastern, and southern India, and its presence has been documented in at least 11 Indian states (3–7).

To the Editor: Scrub typhus, caused by Orientia tsutsugamushi, is one of the most common infectious diseases of rural southern Asia, southeastern Asia, and the western Pacifi c. The disease is transmitted to humans by the bite of larvae of trombiculid mites harboring the pathogen. The disease often appears as a nonspecifi c febrile illness. The clinical picture of scrub typhus is typically associated with fever, rash, myalgia, and diffuse lymphadenopathy (1). Immunofl uorescence assay (IFA) is the test of choice for serodiagnosis of rickettsial diseases (2). Scrub typhus has been reported from northern, eastern, and southern India, and its presence has been documented in at least 11 Indian states (3)(4)(5)(6)(7).
Our study's goal was to retrospectively analyze data of patients with scrub typhus involving the central nervous system. Scrub typhus was suspected on the basis of clinical signs such as febrile illness or fever with rash or eschar. The fever workup profi le (Widal agglutination test, peripheral smear, blood, and urine culture) was noncontributory. Blood samples were obtained after patients gave informed consent. All patients with clinically suspected scrub typhus received antirickettsial drugs (doxycycline and/ or azithromycin) empirically. IFA and PCR of blood samples were performed to confi rm scrub typhus following standard protocol (3). DNA was extracted from the blood sample (buffy coat) by using QIAamp DNA Mini Kit (QIAGEN GmbH, Hilden, Germany) according to the manufacturer's instructions. A standard PCR specifi c for the 56-kDa protein with forward and reverse primers (OtsuF: 5′-AATTGCTAGTGCAATGTCTG-3′ and OtsuR: 5′-GGCATTATAGTAGGC TGAG-3′) was performed (3). PCR products were purifi ed by using the QIAquick PCR Purifi cation Kit (QIA-GEN) according to the manufacturer's instructions. Sequencing reactions were done by using a DNA sequencing kit, dRhodamine Terminator Cycle Sequencing Ready Reaction Mix (Applied Biosystems, Foster City, CA, USA). Sequencing was performed on an ABI PRISM 310 DNA Sequencer (Applied Biosystems). The obtained sequences were identifi ed by comparison with sequences available in Gen-Bank by using the BLAST software (http://blast.ncbi.nlm.nih.gov) (3).
During 2004-2006, scrub typhus was confi rmed in 27 patients; 4 had features of central nervous system involvement. All 4 had fever with altered sensorium and meningeal signs; 2 had seizures. No neurologic focal defi cit was noted, but all showed cerebrospinal fl uid abnormalities. One patient had an eschar, but none had a rash. Serum of 2 patients was subjected to IFA; both samples showed high titers (Table), and PCR for blood was positive for O. tsutsugamushi for all patients. Serum was not subjected to examination for leptospirosis. Patients were treated mainly on the basis of clinical grounds because results of serology were not available immediately. Some clinical features of scrub typhus and leptospirosis are similar, and dual infections have been reported (8); therefore, antimicrobial drugs active against both leptospirosis and scrub typhus were included in treatment regimens. One patient received doxycycline and azithromycin, and the remaining 3 received ceftriaxone in addition to doxycycline. Two patients improved, 1 died, and 1 left hospital against medical advice. The clinical and laboratory details, treatments, and outcomes of all patients are given in the Table. O. tsutsugamushi is an obligate intracellular parasite of professional and nonprofessional phagocytes that invades the central nervous system as part of systemic infection and is found in endothelial cells of blood vessels and in circulating phagocytes. A severe headache occurs almost invariably and has been used as a key clinical criterion for identifying suspected cases. Severe features of central nervous system involvement, such as neck stiffness, neurologic weakness, seizures, delirium, and coma, have been reported. Meningismus or meningitis has been found in 5.7%-13.5% of patients (9). The greatest degree of central nervous system involvement in rickettsial diseases occurs in Rocky Mountain spotted fever and epidemic typhus, followed closely by scrub typhus. The meninges are more commonly involved by O. tsutsugamushi than by other rickettsial infections, and the overall histologic picture in the central nervous system is best described as a meningoencephalitis (9). An exhaustive study of 200 cases of scrub typhus showed central nervous system involvement in most patients. However, focal central nervous system damage was rare, and during the encephalitis stage, few objective neurologic signs were apparent, other than those suggesting more generalized cerebral involvement, such as confusion, tremor, and restlessness (10). Now that it is established that O. tsutsugamushi does invade cerebrospinal fl uid, scrub typhus should be considered a cause of mononuclear meningitis in areas in which it is endemic. In our study 1 patient died despite treatment with doxycycline and azithromycin, suggesting the possibility of resistance to these antimicrobial drugs as recently posited in a study conducted in southern India (6). Scrub typhus in these regions should be further investigated in prospective studies, and clinical isolates should be obtained to evaluate susceptibility to antimicrobial drugs.

Pandemic (H1N1) 2009 and HIV Co-infection
To the Editor: We report a case of pandemic (H1N1) 2009 infection in a man with serologic evidence of HIV-1 infection. The clinical course was complicated by lung and brain involvement (respiratory failure and lethargy), severe leukopenia, and thrombocytopenia, but complications resolved after treatment with oseltamivir (150 mg 2×/d).
In November 2009, a 47-year-old man who had received a diagnosis of hepatitis C infection 8 months earlier sought treatment at Ospedale Santa Maria Nuova, Reggio Emilia, Italy. He had a 3-day history of fever, dry cough, and drowsiness. Eight days before being admitted, the man had resided in the hospital's inpatient detoxifi cation unit, in which at least 10 infl uenzalike cases had been recorded. While in the detoxifi cation unit, he had received methadone, 50 mg 1×/d. Computed tomography images of the brain and radiographs of the chest were normal; ultrasound examination showed upper lobe consolidation of the left lung. Hematochemistry showed high creatine phosphokinase levels, leukocyte count 1,380 cell/mm 3 (reference range 4,000-10,000 cells/mm 3 ), thrombocyte count 34,000 cells/mm 3 (reference range 150,000-450,000 cells/ mm 3 ), partial pressure of oxygen 56 mm Hg, and partial pressure of carbon dioxide 53 mm Hg. Urinalysis results were negative for heroin, cocaine, and alcohol; cerebrospinal fl uid (CSF) analysis results were within normal limits. Thrombocyte count returned to reference range after 2 days, and leukocyte count improved but remained <3,500 cells/mm 3 for 3 weeks. After admission to hospital, the man became lethargic and received noninvasive continuous positive airway pressure ventilation and treated with oseltamivir (150 mg 2×/d for 5 d), as well as with ceftriaxone, and levofl oxacin. Reverse transcription-PCR on a throat swab confi rmed infl uenza subtype H1N1 infection; blood cultures and urine were negative for pneumococcus, and Legionella spp. antigens. In addition, PCR of CSF for enterovirus and herpesvirus had negative results. The patient needed respiratory support for 4 days, after which his mental status and blood gases returned to reference levels. He was discharged from the hospital 2 weeks later.
On day 3 after admission, a nurse was accidentally exposed to the patient's urine through her eye. An ELISA was positive for HIV infection. Negative results for confi rmatory Western blot tests on days 5, 15, and 23 showed the p24 and p41 bands; HIV RNA was >6 million copies/mL, CD4 lymphocytes 51% (reference range 29%-59%). Reverse transcription-PCR for infl uenza subtype H1N1 performed 2 months later on a stored CSF sample gave a negative result; PCR for HIV of the same sample indicated 25,000 copies/mL. In mid-December, because of a further drop in CD4 lymphocytes to 17% (214 cells/mm 3 ) and blood HIV RNA of 2.8 million copies/ mL, the patient started highly active antiretroviral therapy and is being followed up as an outpatient.
Infl uenza (H1N1) and primary HIV infection share many signs and symptoms, such as fever, cough, sore throat, joint or limb pain, and diarrhea. The infections also share uncommon complications of the central nervous system (CNS); e.g., drowsiness, coma, and seizures. We cannot confi rm that CNS involvement in the patient reported here was caused primarily by pandemic (H1N1) 2009, as suggested by infl uenza-like symptoms and the apparent effect of oseltamivir. Nor can we attribute CNS involvement to primary infection with HIV-1 (1); CSF results within normal limits and PCR negative for infl uenza subtype H1N1 do not rule out a causal relationship with pandemic (H1N1) 2009. In fact, the few cases of pandemic (H1N1) 2009 encephalopathy described show similar characteristics among children and adults (2)(3)(4). Alternatively, some authors have attributed HIV in CSF to brain infl ammation and damage (5,6). The severe leukocytopenia and thrombocytopenia in our patient have not been described, even in complicated infl uenza subtype H1N1 infections (7). Because lymphopenia and mild thrombocytopenia are the usual fi ndings, we believe that they probably resulted from HIV-1 or the effect of both viruses.
HIV seroconversion may initially occur during an acute febrile illness resembling infl uenza, and CNS involvement can complicate both infections. During an epidemic, acute HIV infection should also be considered (8). Less frequently, as in the patient described above, the 2 infections can occur simultaneously. History of recent risk behavior for blood exposure and severe leukocytopenia and thrombocytopenia should alert clinicians to other causes and prompt them to offer an HIV test to the patient.