Maculopathy in vision and metamorphopsia in the 1-year clinical outcome and
patients with dengue maculopathy. immune response in these patients.
Demographic details and coexist-
Daniel Hsien-Wen Su* ing conditions are shown in the Table.
To the Editor: A recent article by
and Soon-Phaik Chee*†‡ Patient 1 was a healthcare worker who
Chlebicki et al (1) described 4 patients
*Singapore National Eye Centre, became infected with SARS-associat-
hospitalized for dengue fever who Singapore; †National University of
were found to have retinal hemor- ed coronavirus (CoV) while on duty
Singapore, Singapore; and ‡Singapore
rhages. These patients reported Eye Research Institute, Singapore
with SARS patients. After he was
reduced visual acuity and metamor- transferred to a hospital dedicated to
phopsia, i.e., distorted visual images References SARS management, pulmonary TB
attributable to intrinsic retinal disease was diagnosed (positive acid-fast
1. Chlebicki MP, Ang B, Barkham T, Laude A. bacilli smear on sputum samples).
involving the macula; macular hemor- Retinal hemorrhages in 4 patients with
rhages and exudates were found on dengue fever. Emerg Infect Dis.
Patients 2 and 3 were known to have
retinal examination. The authors con- 2005;11:770–2. cases of pulmonary TB and became
cluded that the retinal hemorrhages 2. Wen KH, Sheu MM, Chung CB, Wang HZ, infected with SARS-CoV after con-
Chen CW. The ocular fundus findings in tact with other patients hospitalized
were responsible for the patients’ dengue fever [article in Chinese]. Gaoxiong
visual symptoms. Yi Xue Ke Xue Za Zhi. 1989;5:24–30.
for SARS. These 2 patients were spu-
This conclusion is misleading 3. Haritoglou C, Scholz F, Bialaslewicz A, tum smear–negative for acid-fast
because retinal hemorrhages alone Klauss V. Ocular manifestations in dengue bacilli, and diagnosis was made on the
fever [article in German]. Ophthalmologe. basis of previous exposure to TB, rel-
cause scotomas. Rather, the accumu- 2000;97:433–6.
lation of subretinal fluid in the macu- 4. Haritoglou C, Dotse SD, Rudolph G,
evant symptoms of typical pulmonary
la results in metamorphopsia and Stephan CM, Thurau SR, Klauss V. A TB, chest radiographs consistent with
blurring of vision. In previous reports tourist with dengue fever and visual loss. active disease, a positive tuberculin
Lancet. 2002;360:1070. skin test result, and the finding of cav-
of patients in whom macular changes 5. Lim WK, Mathur R, Koh A, Yeoh R, Chee
developed from dengue fever, some SP. Ocular manifestations of dengue fever.
ity regression on chest radiographs
were found to have macular hemor- Ophthalmology. 2004;111:2057–64. after anti-TB treatment was initiated.
rhages (2–4). In addition, clinical No cultures were obtained for isola-
examination and investigation of Address for correspondence: Daniel H-W. Su, tion and comparison of Myco-
these patients showed vasculopatho- Singapore National Eye Centre, 11 Third bacterium tuberculosis strains (1). All
logic changes in the macular region Hospital Ave, Singapore 168751; fax: 65-6226- 3 patients had confirmed SARS based
that affected the retinal and choroidal 3395; email: email@example.com on amplification of SARS-CoV RNA
blood vessels (5), although the tissues by reverse transcriptase–polymerase
of the periphery tended to be spared. chain reaction (RT-PCR) from sputum
A fluorescein angiograph of the retina and stool specimens (2). Patients 2
showed knobby hyperfluorescence of and 3 recovered without complica-
the retinal arterioles with minimal tions; patient 1 had the most severe
leakage, as well as some spots of leak- disease and required mechanical ven-
age at the level of the retinal pigment tilation in an intensive care unit before
epithelium. An indocyanine green recovering.
angiograph showed diffuse hyperfluo- Both cellular and humoral immu-
rescence of the choroid. These patho- Tuberculosis and nity were evaluated during the follow-
logic changes in the macula were the SARS, China up of these patients. T-lymphocyte
most likely cause of the blurring of subsets were measured 6 months after
vision in such patients, which has To the Editor: As part of a cohort disease onset by flow cytometry using
been the case in our experience. study of 83 patients with severe acute fluorescein isothiocyanate–labeled
The article by Chlebicki et al. did respiratory syndrome (SARS) in specific monoclonal antibodies.
not state whether these procedures Beijing, China, we conducted a fol- Compared to other SARS patients (n
had been performed on their patients low-up study of all the patients by = 47), the 3 patients with TB had
to confirm or exclude retinal or routine medical examination. During lower mean CD4+ T cells (368.4/µL
choroidal vasculopathy in the macula. the process, 3 patients with chest radi- vs. 656.6/µL, respectively; p = 0.05)
Therefore, these authors would have ographs consistent with active disease and lower mean CD8+ T cells
had difficulty concluding that retinal were identified as having pulmonary (371.0/µL vs. 490.1/µL, respectively;
hemorrhages caused blurring of tuberculosis (TB). Here we describe p = 0.39). SARS-CoV immunoglobu-
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006 707
lin G (IgG) antibody titers were meas- SARS (6) and corticoid therapy. Such AI51915), and the National 863 Program
ured by enzyme-linked immunosor- phenomena have been described with of China (2003AA208406, 2003AA20
bent assay kit (Huada Company, other viral infections such as measles 8412C).
Beijing, China) at months 1, 2, 3, 4, 7, and HIV (7,8). By contrast, patients 2
10, and 16 after disease onset. (Titers and 3 were known TB patients who Wei Liu,* Arnaud Fontanet,†
were not measured for the 3 TB acquired SARS through exposure to Pan-He Zhang,* Lin Zhan,*
patients at month l.) Compared to SARS patients in the same hospital Zhong-Tao Xin,‡ Fang Tang,*
most (26 [78.8%] of 33) other SARS wards. Both diseases are known to be Laurence Baril,†
patients whose antibodies remained transiently immunosuppressive (6,9), and Wu-Chun Cao*
detectable throughout follow-up, 2 of and their combined effect resulted in *Beijing Institute of Microbiology and
the 3 TB patients (patients 1 and 3) more pronounced CD4+ cell decreas- Epidemiology, Beijing, People’s Republic of
had undetectable antibody titers as of es in coinfected SARS patients than China, †Institut Pasteur, Paris, France; and
months 7 and 16, respectively. In others. Such immunosuppression also ‡Beijing Institute of Basic Medical
Sciences, Beijing, People’s Republic of
patient 1, antibody titers, when resulted in poorer IgG antibody
detectable, were unusually low (40). response in coinfected SARS patients
Both patients 1 and 3 had prolonged than in others and delayed viral clear-
viral excretion in stools, sputum, or ance, as shown by longer viral excre-
both. While the median (range) dura- tion in sputum and stools. While viral 1. Crawford JT, Braden CR, Schable BA,
Onorato ID. National Tuberculosis
tion of virus excretion in stools and excretion could be prolonged in coin-
Genotyping and Surveillance Network:
sputa for the entire measurable cohort fected patients, no virus could be iso- design and methods. Emerg Infect Dis.
(n = 56) was 27 (16–127) and 21 lated from any RT-PCR–positive 2002;8:1192–6.
(14–52) days, respectively (3), it was specimen collected after 6 weeks of 2. Revised U.S. surveillance case definition
for severe acute respiratory syndrome
125 and 16 days for patient 1, and 109 illness, which suggests that excreted
(SARS) and update on SARS cases—
and 52 days for patient 3 (viral excre- viruses were no longer infectious (3). United States and worldwide, December
tion data could not be obtained from These case reports remind us of the 2003. MMWR Morb Mortal Wkly Rep.
patient 2 because sequential speci- importance of strict isolation of SARS 2003;52:1202–6.
3. Liu W, Tang F, Fontanet A, Zhan L, Zhao
mens for detection were unavailable). patients, careful use of steroids for
QM, Zhang PH, et al. Long-term SARS-
TB in SARS patients has been their case management, and the possi- coronavirus excretion from a patient cohort
reported on rare occasions (4,5). In a bility of coinfection with TB in SARS in China. Emerg Infect Dis.
cohort of 236 patients in Singapore, it patients with incomplete recovery. 2004;10:1841–3.
4. Low JGH, Lee CC, Leo YS. Severe acute
was diagnosed in 2 patients after
respiratory syndrome and pulmonary tuber-
recovery from SARS (4). As with culosis. Clin Infect Dis. 2004;38:e123–5.
This work was partly supported by
patient 1 in this study, TB had devel- 5. Centers for Disease Control and
the Programme de Recherche en Réseaux Prevention. Nosocomial transmission of
oped after the patient acquired SARS,
Franco-Chinois (P2R), the EC grant EPIS- Mycobacterium tuberculosis found through
most likely as the result of reactiva-
ARS (SP22-CT-2004-511063, SP22-CT- screening for severe acute respiratory syn-
tion of past infection or new infection drome—Taipei, Taiwan, 2003. MMWR
2004-003824), the National Institutes of
with M. tuberculosis, while temporar- Morb Mortal Wkly Rep. 2004;53:321–2.
Health CIPRA Project (NIH U19
ily immunosuppressed because of
708 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006
6. Li T, Qiu Z, Zhang L, Han Y, He W, Liu Z, January 1, 2003, by adapting the Two patients died (case fatality
et al. Significant changes of peripheral T existing enhanced tetanus surveil- 8%). Of 23 survivors, 2 had mild dis-
lymphocyte subsets in patients with severe
acute respiratory syndrome. J Infect Dis. lance. A case was defined as mild-to- ease and 21 required intensive treat-
2004;189:648–51. moderate trismus and at least 1 of the ment for a median of 40 days (range
7. Griffin DE, Bellini WJ. Measles. In: Fields following: spasticity, dysphagia, res- 24–65 in 15 cases with complete infor-
BN, editor. Virology. New York: Raven piratory embarrassment, spasms, mation). Tetanus immunization status
Press; 1996. p. 1267–312.
8. Havlir DV, Barnes PF. Current concepts: autonomic dysfunction, in a person available for 20 case-patients (based
tuberculosis in patients with human who injected drugs in the month on medical records or patient and
immunodeficiency virus infection. N Engl J before symptom onset. parental recall) indicated that only 1
Med. 1999;340:367–73. Twenty-five cases were reported patient (with severe disease) had
9. Frieden TR, Sterling TR, Munsiff SS, Watt
CJ, Dye C. Tuberculosis. Lancet. from July 2003 to September 2004 received the 5 doses necessary for
2003;362:887–99. (Figure). Thirteen (50%) were complete coverage. Nine patients were
women; the median age of male and never vaccinated. Twelve of 14
Address for correspondence: Wu-Chun Cao, female patients was 39 and 32 years patients tested for tetanus immunity on
Beijing Institute of Microbiology and of age, respectively (range 20–53, p = admission by a standard indirect
Epidemiology, State Key Laboratory of 0.1). Twenty patients were white, and enzyme-linked immunosorbent assay
Pathogen and Biosecurity, Beijing, People’s 1 was Chinese (information was miss- had antibody levels lower than the cut-
Republic of China; fax: 86-10-6389-6082; ing for 4). None reported travel over- off value for protection (<0.1 IU/mL).
email: firstname.lastname@example.org seas before becoming sick. Seventeen One patient with severe disease had a
of 21 patients with information level just above the cutoff value and 1
reported having injected heroin intra- patient with mild disease had a protec-
muscularly or subcutaneously (pop- tive antibody level. Clostridium tetani
ping) or having missed veins. Most was isolated from 2 patients; tetanus
patients (16/25) came to the hospital toxin was detected in serum from 1
with severe generalized tetanus. and also from another patient. Other
Injection site infections were common anaerobes, including C. novyi, C. his-
Tetanus in (17/19). tolyticum, and C. perfringens, were
Injecting Drug Figure. Cases of tetanus in injecting
drug users by residence (25 cases) and
place from which heroin was supplied
Kingdom (14 cases with information), United
Kingdom, July 2003–September 2004.
The large circle indicates the Liverpool
To the Editor: The epidemiology
area. Squares indicate the residence of
of tetanus in the United Kingdom patients for whom the origin of heroin
changed in 2003 when a cluster of was not reported, open circles indicate
cases in injecting drug users (IDUs) the residence of patients for whom the
occurred (1,2). Before 2003, the inci- origin of heroin was reported, and solid
circles indicate the origin of heroin.
dence of tetanus was low in the
United Kingdom, with occasional
cases predominantly in unvaccinated
elderly persons (3). The situation con-
trasted with the United States where
injecting drug use is commonly
reported among persons with tetanus
We investigated the UK cluster to
identify the source of infection and
opportunities for prevention. We
ascertained cases through statutory
and nonstatutory reporting to the
Health Protection Agency and collect-
ed additional information on IDUs for
all reported cases of tetanus since
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006 709