Antiviral activity of cidofovir and ribavirin against the new human adenovirus subtype 14a that is associated with severe pneumonia.

728 • CID 2008:47 (1 September) • CORRESPONDENCE
1 S E P T E M B E R
Correspondence
Table 1. Characteristics of 61 patients with tropical spastic paraparesis/human T lymphotropic virus type 1 (HTLV-1)–associated
myelopathy (TSP/HAM), by sex.
Variables
Female
patients
(n p 42)
Male
patients
(n p 19)
All
patients
(n p 61)
Age, mean years � SD (range) 49 � 11 (23–74) 41 � 12 (24–72) 49 � 12 (23–74)
Time from symptom onset to diagnosis of TSP/HAM, years 9 � 1 9 � 1 9 � 1
Age of TSP/HAM diagnosis, years 49 � 12 50 � 15 49 � 13
Age of onset of TSP/HAM symptoms, years 40 � 11 41 � 18 41 � 13
HTLV-1 load, copies/104 PBMCs 889 � 1668 1117 � 1580 950 � 1595
NOTE. Data are mean , unless otherwise indicated. Only 3 men received a diagnosis of TSP/HAM when they were 160 years of age.value�SD
Low Risk of Tropical Spastic
Paraparesis/Human T
Lymphotropic Virus Type
1–Associated Myelopathy
Development among Persons
Aged 150 Years
To the Editor—Tropical spastic para-
paresis/human T lymphotropic virus type
1 (HTLV-1)–associated myelopathy (TSP/
HAM) is an immune-mediated disease [1]
characterized by chronic, slowly progres-
sive, spastic paraparesis with bladder dis-
turbances, absent or mild sensory loss, and
lower back pain in the absence of spinal
cord compression, as well as by seropos-
itivity for HTLV-1 antibodies [2]. Al-
though the usual presentation is charac-
terized by a slow progression, 21.5% of
the patients may experience a rapid pro-
gression, with severe disability developing
2 years after the onset of symptoms [3].
The estimated incidence of TSP/HAM
ranges from 0.25% to 1% after 30–40 years
of incubation [4]. The mean age of onset
of disease is 40 years, and the disease oc-
curs predominantly in women [5]. Several
factors have been ascribed as effecting clin-
ical outcome, including high HTLV-1 load
[6], genetic background [7], route of
transmission (i.e., breastfeeding or trans-
fusion), and high antibody titers [8]. Be-
cause of the range of clinical evolution of
progression, we decided to assess the time
of onset of signs and/or symptoms of TSP/
HAM in our HTLV Service in Sao Paulo,
Brazil.
Over the past 10 years, our research
group has observed 360 HTLV-1–infected
patients, 61 of whom have received a di-
agnosis of TSP/HAM. During the follow-
up period, 2 additional patients (among
299 healthy asymptomatic carriers) de-
veloped TSP/HAM (data not shown), with
diagnosis based on revised Kagoshima cri-
teria [9]. Both patients were !30 years of
age, and their parents had also received
diagnoses of TSP/HAM. Overall, the mean
age of TSP/HAM development was 47
years, and the mean age at onset of symp-
toms was 40 years. None of the asymp-
tomatic carriers aged 150 years developed
symptoms during the follow-up period.
Although 3 men received a diagnosis when
they were 160 years of age, initial symp-
toms primarily occurred in individuals
aged !40 years (table 1). HTLV-1 DNA
loads were similar, regardless of sex and
age.
In contrast to adult T cell leukemia, the
incidence of TSP/HAM decreases with
age. Thus, the progression of TSP/HAM
is similar to that of multiple sclerosis; for
both conditions, few cases have been re-
ported in persons aged 160 years [10]. A
possible explanation for this is that, be-
cause TSP/HAM is an immune-mediated
illness, lack of CD8+ cell hyperactivity in
older patients decreases the risk of clinical
development [11]. This observation may
have implications for counseling and
medical practices in areas where HTLV-1
infection is endemic.
Acknowledgments
Potential conflicts of interest. J.C. and
A.C.P.O.: no conflicts.
Jorge Casseb and Augusto C. P. Oliveira
Human T Lymphotropic Virus Clinic,
Institute of Infectious Diseases “Emı´lio Ribas,”
Sa˜o Paulo, Brazil
References
1. Iwasaki Y. Pathology of chronic myelopathy
associated with HTLV-1 infection (HAM/
TSP). J Neurol Sci 1990; 96:103–23.
2. Osame M, Usuku K, Izumo S, et al. HTLV-I
associated myelopathy, a new clinical entity.
Lancet 1986; 1:1031–2.
3. Matsuzaki T, Nakagawa M, Nagai M, et al.
HTLV-I proviral load correlates with progres-
sion of motor disability in HAM/TSP: analysis
of 239 HAM/TSP patients including 64 pa-
tients followed up for 10 years. J Neurovirol
2001; 7:228–34.
4. Kaplan JE, Osame M, Kubota H, et al. The
risk of development of HTLV-I–associated
myelopathy/tropical spastic paraparesis
among persons infected with HTLV-I. J Acquir
Immune Defic Syndr 1990; 3:1096–101.
5. Gotuzzo E, Cabrera J, Deza L, et al. Clinical
characteristics of patients in Peru with human
T cell lymphotropic virus type 1–associated

CORRESPONDENCE • CID 2008:47 (1 September) • 729
tropical spastic paraparesis. Clin Infect Dis
2004; 39:939–44.
6. Olindo S, Le´zin A, Cabre P, et al. HTLV-1
proviral load in peripheral blood mononu-
clear cells quantified in 100 HAM/TSP pa-
tients: a marker of disease progression. J Neu-
rol Sci 2005; 237:53–9.
7. Jeffery KJ, Usuku K, Hall SE, et al. HLA alleles
determine human T-lymphotropic virus-I
(HTLV-I) proviral load and the risk of HTLV-
I–associated myelopathy. Proc Natl Acad Sci
U S A 1999; 96:3848–53.
8. Nakagawa M, Izumo S, Ijichi S, et al. HTLV-
I–associated myelopathy: analysis of 213 pa-
tients based on clinical features and laboratory
findings. J Neurovirol 1995; 1:50–61.
9. De Castro-Costa CM, Arau´jo AQ, Barreto
MM, et al. Proposal for diagnostic criteria of
tropical spastic paraparesis/HTLV-I-associated
myelopathy (TSP/HAM). AIDS Res Hum Re-
troviruses 2006; 22:931–5.
10. Ebers GC. Prognostic factors for multiple scle-
rosis: the importance of natural history stud-
ies. J Neurol 2005; 252(Suppl 3):iii, 15–20.
11. Effros RB, Dagarag M, Spaulding C, Man J.
The role of CD8+ T-cell replicative senescence
in human aging. Immunol Rev 2005; 205:
147–57.
Reprints or correspondence: Dr. Jorge Casseb, Institute of
Tropical Medicine, Sa˜o Paulo University, Ave. Dr. Eneas de
Carvalho Aguiar 500, Bldg. II, 3rd flr., Sa˜o Paulo, SP Brazil
05403–907 (Jorge_casseb@yahoo.com.br).
Clinical Infectious Diseases 2008; 47:728–9
� 2008 by the Infectious Diseases Society of America. All
rights reserved. 1058-4838/2008/4705-0022$15.00
DOI: 10.1086/590969
Pneumococcal Necrotizing
Pneumonia and Pleural Fluid
Lactate Dehydrogenase Level
To the Editor—I read with interest the
recent article by Bender et al. [1] about
pneumococcal necrotizing pneumonia
(NP) in children. The results of the study
confirm the severity of this entity in chil-
dren (as reported elsewhere [2–4]), and
the study enhances the previous findings
with more detailed serotype analysis. Sim-
ilar to the findings of Bender et al. [1], we
have recently seen at our institution severe
presentations and bad outcomes of pneu-
mococcal NP associated with serotype 3
and other serotypes [5, 6]. Of concern to
us, very high initial lactate dehydrogenase
(LDH) levels were seen in pleural fluid
specimens obtained from children who re-
ceived an initial diagnosis of pneumococ-
cal empyema and who subsequently de-
veloped severe NP and required extensive
decortications, segmentectomies, or lo-
bectomies. In retrospect, some of these
children eventually would have benefited
from earlier surgical interventions.
Although children with complicated
pneumococcal pneumonia who require
decortication drainage may have higher
levels of LDH in pleural fluid [7], it is
uncertain whether patients with pneu-
mococcal NP exhibit significantly higher
LDH levels in pleural fluid, compared with
patients with NP due to Staphylococcus au-
reus or other organisms. In the 2 largest
series of NP in children, an in-depth anal-
ysis of this issue was not performed [8, 9].
Similarly, studies comparing pleural fluid
LDH levels in children with complicated
pneumococcal pneumonia caused by
non-pneumococcal conjugate vaccine–7
serotypes versus pneumococcal conjugate
vaccine–7 serotypes have not been
performed.
LDH, a useful parameter in pleural fluid
analysis for patients with complicated
pneumonia (among others), reflects cel-
lular injury and can be released by cells
undergoing either primary or secondary
necrosis. Recent research findings in mice
have revealed that, during severe pulmo-
nary inflammation (i.e., bacterial pneu-
monia), apoptotic neutrophils undergoing
secondary necrosis are the primary source
of LDH in bronchoalveolar lavage fluid
[10]. The clinical relevance of these find-
ings is unknown for children with com-
plicated pneumonia; however, it is likely
that these changes also occur in the lungs
of individuals with NP, which may par-
tially explain the elevated LDH levels seen
in these individuals. These high or ab-
normally high values should alert clini-
cians about the possibility of ongoing ne-
crosis or liquefaction of the pulmonary
parenchyma, and clinicians must give spe-
cial attention to this issue in the present
era, in which NP is more commonly
reported.
The study by Bender et al. [1] is ad-
mirable, and I understand that the objec-
tives that they intended at the beginning
were different from the issues that I discuss
here. However, considering that their
study involved children with necrotizing
pneumonia, mention of other aspects of
this entity is important and would benefit
their study. No mention was made about
pleural fluid LDH levels or about findings
of examination of lung biopsy specimens
from those who underwent surgical pro-
cedures and the respective intraoperative
findings. Could the authors provide LDH
findings in this cohort (i.e., ranges and
mean values)? Were there any serotypes
associated with higher levels of LDH?
Acknowledgments
Potential conflicts of interest. R.U.-G. has re-
ceived funding from Wyeth for traveling to pneu-
mococcus-related conferences to present posters
and give lectures.
Rolando Ulloa-Gutierrez
Servicio de Infectologı´a Pedia´trica,
Hospital Nacional de Nin˜os de Costa Rica,
“Dr. Carlos Sa´enz Herrera,”
San Jose´, Costa Rica
References
1. Bender JM, Ampofo K, Korgenski K, et al.
Pneumococcal necrotizing pneumonia in
Utah: does serotype matter? Clin Infect Dis
2008; 46:1346–52.
2. McCarthy VP, Patamasucon P, Gaines T, Lucas
MA. Necrotizing pneumococcal pneumonia
in childhood. Pediatr Pulmonol 1999; 28:
217–21.
3. Hsieh YC, Hsueh PR, Lu CY, Lee PI, Lee CY,
Huang LM. Clinical manifestations and mo-
lecular epidemiology of necrotizing pneu-
monia and empyema caused by Streptococcus
pneumoniae in children in Taiwan. Clin Infect
Dis 2004; 38:830–5.
4. Hsieh YC, Hsiao CH, Tsao PN, et al. Necro-
tizing pneumococcal pneumonia in children:
the role of pulmonary gangrene. Pediatr Pul-
monol 2006; 41:623–9.
5. Soto-Martı´nez M, Campos-Goussen C, Sori-
ano-Fallas A, Aco´n-Rojas F, Soto-Quiro´s M,
Ulloa-Gutierrez R. Pneumococcal pleural em-
pyema in Costa Rican children: 1995–2007
[abstract 201]. In: Program and abstracts of
the 45th Annual Meeting of the Infectious Dis-
eases Society of America (San Diego). 2007.
6. Ulloa-Gutierrez R, Soto-Martı´nez M, Cama-
cho-Badilla K, Soriano-Fallas A, Soto-Quiro´s
M. Pneumococcal necrotizing pneumonia
(NP): are we giving enough attention to lactate
dehydrogenase values [abstract P1-084]? In:
Program and abstracts of the 6th International
Symposium on Pneumococci and Pneumo-
coccal Diseases (Reykjavik). 2008.

730 • CID 2008:47 (1 September) • CORRESPONDENCE
7. Tan TQ, Mason EO Jr, Wald ER, et al. Clinical
characteristics of children with complicated
pneumonia caused by Streptococcus pneumon-
iae. Pediatrics 2002; 110:1–6.
8. Hacimustafaoglu M, Celebi S, Sarimehmet H,
Gurpinar A, Ercan I. Necrotizing pneumonia
in children. Acta Paediatr 2004; 93:1172–7.
9. Sawicki GS, Lu FL, Valim C, Cleveland RH,
Colin AA. Necrotizing pneumonia is an in-
creasingly detected complication of pneu-
monia in children. Eur Respir J 2008; 31:
1285–91.
10. Rydell-To¨rma¨nen K, Uller L, Erjefa¨lt JS. Direct
evidence of secondary necrosis of neutrophils
during intense lung inflammation. Eur Respir
J 2006; 28:268–74.
Reprints or correspondence: Dr. Rolando Ulloa-Gutierrez, Ser-
vicio de Infectologı´a Pedia´trica, Hospital Nacional de Nin˜os
de Costa Rica, “Dr. Carlos Sa´enz Herrera,” PO Box 1654–
1000, Ave. Paseo Colo´n, San Jose´, Costa Rica (rolandoug
@racsa.co.cr).
Clinical Infectious Diseases 2008; 47:729–30
� 2008 by the Infectious Diseases Society of America. All
rights reserved. 1058-4838/2008/4705-0023$15.00
DOI: 10.1086/590974
Reply to Ulloa-Gutierrez
To the Editor—We thank Dr. Ulloa-
Gutierrez [1] for his interest in our recent
article about the association of Streptococ-
cus pneumoniae serotypes with necrotizing
pneumonia in children in Utah [2]. Ulloa-
Gutierrez [1] addressed parameters other
than S. pneumoniae serotype that may be
predictors of severity. Specifically, an ele-
vated lactate dehydrogenase (LDH) level
appears to be associated with severe nec-
rotizing pneumonia caused by serotype 3
in the Hospital Nacional de Nin˜os de
Costa Rica (San Jose´) [1]. An elevated
LDH level likely represents cellular dam-
age to lung parenchyma, and it is reason-
able to hypothesize that the degree of el-
evation correlates with the extent of
necrosis. If S. pneumoniae serotype 3
causes more extensive necrosis, it stands
to reason that LDH levels will be more
elevated in patients with pneumonia
caused by serotype 3.
At the suggestion of Ulloa-Gutierrez
[1], we retrospectively evaluated the pleu-
ral fluid LDH levels in the 14 patients with
serotype 3 pneumonia in our study. Five
of the 14 patients had pleural fluid LDH
levels measured. The median LDH level
was 49,460 U/L (range, 39,700–118,041 U/
L). These values represent a 40–120-fold
increase over the upper limit of normal
for serum LDH level in our laboratory
(975 U/L). These findings appear to be
consistent with those of Ulloa-Gutierrez
[1]. However, care should be taken when
comparing pleural fluid indices from ret-
rospective studies because of the variability
in time to presentation and other clinical
and laboratory factors.
Complicated pneumonia in children is
a growing problem worldwide. We agree
with Ulloa-Gutierrez [1] that pleural fluid
indices might prove to be a useful adjunct
in predicting outcomes and directing
management. In a separate study, we dem-
onstrated that among children with em-
pyema, those with pleural fluid indices in-
cluding higher WBC counts, lower glucose
values, and the presence of bacteria on
Gram stain or culture were more likely to
require surgical intervention than were
those without these findings [3]. The de-
gree of LDH elevation, however, was not
associated with the need for surgical
intervention.
Acknowledgments
Potential conflicts of interest. All authors: no
conflicts.
Jeffrey M. Bender, Andrew T. Pavia,
and Carrie L. Byington
Division of Infectious Diseases, Department of
Pediatrics, University of Utah, Salt Lake City
References
1. Ulloa-Gutierrez R. Pnuemococcal necrotizing
pneumonia and pleural fluid lactate dehydro-
genase level. Clin Infect Dis 2008; 47:729–30 (in
this issue).
2. Bender JM, Ampofo K, Korgenski K, et al.
Pneumococcal necrotizing pneumonia in Utah:
does serotype matter? Clin Infect Dis 2008; 46:
1346–52.
3. Ampofo K, Byington C, Meyers R, et al. Pre-
dictors of need for salvage therapy with video-
assisted thoracoscopy surgery after chest tube
and fibrinolysis in the management of para-
pneumonic empyema in children [abstract
3532.206]. In: Program and abstracts of the
Pediatric Academic Societies Annual Meeting
(San Francisco). 2006.
Reprints or correspondence: Dr. Jeffrey M. Bender, University
of Utah, Div. of Pediatric Infectious Diseases, Dept. of Pe-
diatrics, 30 N. 1900 East, Salt Lake City, Utah 84132
(Jeffrey.Bender@hsc.utah.edu).
Clinical Infectious Diseases 2008; 47:730
� 2008 by the Infectious Diseases Society of America. All
rights reserved. 1058-4838/2008/4705-0024$15.00
DOI: 10.1086/590975
Natural Products
and Wound Management:
A Never-Ending Story
To the Editor—Kwakman et al. [1] eval-
uated the efficacy of Revamil medical-
grade honey as a topical antimicrobial
agent for prevention or treatment of in-
fection, including those caused by multi-
drug-resistant bacteria, and they proposed
this natural product for the treatment of
wound infections. Their article raises some
questions about the use of natural prod-
ucts in modern medicine and about what
can be learned from the medical practices
of ancient centuries [1].
In some ways, the history of medicine
is strictly bound to the care of wounds,
especially those injuries caused in battles.
In Mesopotamia (2100 bc), after cleaning
a wound with beer, a bandage with wine
and turpentine was applied. In Ancient
Egypt, a natural product, such as honey,
became common in wound management:
after irrigating a wound with wine, phy-
sicians would then cover the wound with
fat and honey. In Greece, during the time
of Homer, wounds were covered and
treated with herbs, whereas infected
wounds were healed by dropping scrap-
ings from bronze spears into them.
Four hundred years later, Hippocrates
taught that wounds should be washed with
wine, bandaged, and then saturated with
more wine. Galen, of ancient Rome,
cleaned wounds with vinegar but nev-
ertheless believed that suppuration (“pus
bonum et laudabile”) was necessary for
healing. This belief prevailed for many
centuries.
When gunpowder was first used in bat-
tle, boiling oil was used as wound treat-
ment, with the aim “to inactivate” the
gunpowder effect within gunshot wounds.

CORRESPONDENCE • CID 2008:47 (1 September) • 731
Table 1. Antiviral activity of ribavirin and cidofovir against human adenovirus type
5 of species C (HAdV-C5) and human adenovirus subtype 14a (HAdV-B14a) in a lung
cancer cell line (A549 cells).
Species
Ribavirin Cidofovir
IC50 IC90 IC50 IC90
HAdV-B14a 2.6 (1.9–3.7) 734.3 (388.0–1557.4) 0.8 (0.7–0.9) 16.0 (12.1–21.9)
HAdV-C5 1.7 (1.1–2.8) 29.0 (15.1–64.4) 0.4 (0.3–0.5) 19.4 (12.1–33.9)
NOTE. Data are expressed as mean mM (95% CI). IC50, 50% inhibitory concentration; IC90, 90%
inhibitory concentration.
This practice was abandoned when, in
1542, Ambroise Pare´ performed one of the
first cohort studies to demonstrate that
egg yolk and turpentine were better than
boiling oil for treatment of gunshot
wounds.
There are limited data on natural prod-
ucts, such as yogurt (which seems to help
in controlling wound odor [2]), tea tree
oil, and potato peelings; however, there is
a great interest in the use of honey for
wound healing [3]. Indeed, honey kills
staphylococci [4], including the fearsome
community-acquired methicillin-resistant
Staphylococcus aureus [5], within a few
hours; it has anti-inflammatory activity
[6]; and its hypertonicity provides anti-
septic activity. However, clinical data on
the effect of honey on wound manage-
ment are controversial [7, 8], often be-
cause of the low quality of some studies
[9] and perhaps because of different an-
timicrobial activities of different honey
types [10].
Although science and technology
march ever forward, we must not fail
to exploit the potential of natural prod-
ucts, because they may represent—as in
ancient times—a hope for modern
medicine.
Acknowledgments
Potential conflicts of interest. N.P.: no
conflicts.
Nicola Petrosillo
2nd Infectious Diseases Division,
National Institute for Infectious Diseases,
“L. Spallanzani,” Rome, Italy
References
1. Kwakman PHS, Van den Akker JPC, Guclu
A, et al. Medical-grade honey kills antibiotic-
resistant bacteria in vitro and eradicates skin
colonization. Clin Infect Dis 2008; 46:
1677–82.
2. Schulte MJ. Yogurt helps to control wound
odor. Oncol Nurs Forum 1993; 20:1262.
3. Molan PC. The evidence supporting the use
of honey as a wound dressing. Int J Low Ex-
trem Wounds 2006; 5:40–54.
4. French VM, Cooper RA, Molan PC. The an-
timicrobial activity of honey against coagulase-
negative staphylococci. J Antimicrob Che-
mother 2005; 56:228–31.
5. Maeda Y, Loughrey A, Earle JA et al. Anti-
bacterial activity of honey against community-
associated methicillin-resistant Staphylococcus
aureus (CA-MRSA). Complement Ther Clin
Pract 2008; 14:77–82.
6. Tonks AJ, Cooper RA, Jones KP, Blair S, Par-
ton J, Tonks A. Honey stimulates inflamma-
tory cytokine production from monocytes.
Cytokine 2003; 21:242–7.
7. Jull A, Walker N, Paraq V, et al. Randomized
clinical trial of honey-impregnated dressings
for venous leg ulcers. Br J Surg 2008; 95:
175–82.
8. Mphande AN, Killowe C, Phalira S, Jones HW,
Harrison WJ. Effects of honey and sugar dress-
ings on wound healing. J Wound Care
2007; 16:317–9.
9. Moore OA, Smith LA, Campbell F, Seers K,
McQuay HJ, Moore RA. Systematic review of
the use of honey as a wound dressing. BMC
Complement Altern Med 2001; 1:2.
10. Karpelowsky J, Allsopp M. Wound healing
with honey—a randomized controlled trial. S
Afr Med J 2007; 97:314, 316.
Reprints or correspondence: Dr. Nicola Petrosillo, 2nd Infec-
tious Diseases Division, National Institute for Infectious Dis-
eases, “L. Spallanzani,” Via Portuense, 292–00149 Rome, Italy
(petrosillo@inmi.it).
Clinical Infectious Diseases 2008; 47:730–1
� 2008 by the Infectious Diseases Society of America. All
rights reserved. 1058-4838/2008/4705-0025$15.00
DOI: 10.1086/590972
Antiviral Activity of
Cidofovir and Ribavirin
against the New Human
Adenovirus Subtype 14a
That Is Associated with
Severe Pneumonia
To the Editor—Louie et al. [1] have de-
scribed 3 cases of severe pneumonia
caused by a new pneumotropic human
adenovirus subtype 14a (HAdV-B14a).
Moreover, several outbreaks of HAdV-
B14a infection with a high mortality have
been described in the United States, sug-
gesting emergence of a highly pathogenic
virus [2]. The emergence of HAdV-B14a
as a respiratory pathogen has also been
observed at US military training centers
[3].
An experimental antiviral therapy with
cidofovir for a single case of HAdV-B14a
pneumonia was reported by Louie et al.
[1], although it achieved limited clinical
success. HAdV-B14a is a member of spe-
cies B (subspecies B2) human adenovi-
ruses. In vitro susceptibility of human ad-
enoviruses to antiviral drugs is species
dependent. Human adenoviruses of spe-
cies C have been shown to be susceptible
in vitro to both ribavirin and cidofovir,
whereas the susceptibility of human ade-
noviruses of other species to ribavirin has
been shown to be significantly lower [4,
5]. Nevertheless, testing the in vitro sus-
ceptibility of HAdV-B14a to ribavirin
seems to be reasonable, because high con-
centrations of ribavirin (11000 mM) can
be achieved in respiratory secretions with
use of high-dose aerosol therapy, which
has been developed for treatment of re-
spiratory syncytial virus infection [6].
Therefore, we tested the antiviral activ-
ity of cidofovir and ribavirin against
HAdV-B14a (Portland isolate, generously
provided by D. Erdmann, Centers for Dis-
ease Control and Prevention). A rapid
quantitative PCR-based assay for testing
antiviral agents against human adenovi-
ruses on a lung cancer cell line (A549) was
used as described elsewhere [5]. Human
adenovirus type 5 of species C (HAdV-
C5) served as a positive control. In vitro
susceptibility of HAdV-B14a to cidofovir
was high and in the same range as that of
in vitro susceptibility of HAdV-C5. Peak
plasma levels reported after intravenous

732 • CID 2008:47 (1 September) • CORRESPONDENCE
application of 5 mg of cidofovir per ki-
logram of body weight were ∼5-fold
higher than an inhibitory concentration of
90% (IC90) against HAdV-B14a (table 1)
[7]. Thus, cidofovir holds promise to be
active against HAdV-B14a replication in
vivo. In the case of ribavirin, the low 50%
inhibitory concentration (IC50) value sug-
gested an in vitro susceptibility of HAdV-
B14a that is similar to that of HAdV-C5
(table 1). However, a 90% inhibition of
HAdV-B14a replication was not achieved
with noncytotoxic concentrations (table 1;
50% cytotoxic concentration of ribavirin,
802 mM). For comparison, 90% inhibi-
tions of HAdV-C5 and HAdV-B11, which
is closely related to HAdV-B14a, were
achieved with noncytotoxic ribavirin con-
centrations of 29 mM and 38 mM, respec-
tively [5]. Therefore, the interpretation of
HAdV-B14a in vitro susceptibility data re-
quires caution, in spite of very high con-
centrations (11000 mM ribavirin) achieved
in respiratory secretions by high-dose aer-
osol therapy.
In conclusion, intravenous application
of cidofovir should reach a plasma con-
centration well above in vitro IC50 and IC90
values for HAdV-14a. Although lung con-
centrations may be lower and early initi-
ation of therapy may be crucial, cidofovir
holds promise as an antiviral agent for the
treatment of severe lower respiratory tract
infections caused by HAdV-B14a. By con-
trast, high IC90 values of ribavirin in the
range of cytotoxic concentrations may
lead to clinical failure, although high al-
veolar concentrations can be achieved us-
ing aerosolized ribavirin.
Acknowledgments
Potential conflicts of interest. All authors: no
conflicts.
Sebastian Darr, Ijad Madisch,
and Albert Heim
Institut fu¨r Virologie, Medizinische Hochschule
Hannover, Hannover, Germany
References
1. Louie JK, Kajon AE, Holodniy M, et al. Severe
pneumonia due to adenovirus serotype 14: a
new respiratory threat? Clin Infect Dis 2008;
46:421–5.
2. Centers for Disease Control and Prevention.
Acute respiratory disease associated with ade-
novirus serotype 14—four states, 2006–2007.
MMWR Morb Mortal Wkly Rep 2007; 56:
1181–4.
3. Metzgar D, Osuna M, Kajon AE, Hawksworth
AW, Irvine M, Russell KL. Abrupt emergence
of diverse species B adenoviruses at US military
recruit training centers. J Infect Dis 2007; 196:
1465–73.
4. Morfin F, Dupuis-Girod S, Mundweiler S, et al.
In vitro susceptibility of adenovirus to antiviral
drugs is species-dependent. Antivir Ther
2005; 10:225–9.
5. Stock R, Harste G, Madisch I, Heim A. A rapid
quantitative PCR-based assay for testing anti-
viral agents against human adenoviruses dem-
onstrates type specific differences in ribavirin
activity. Antiviral Res 2006; 72:34–41.
6. Englund JA, Piedra PA, Jefferson LS, Wilson
SZ, Taber LH, Gilbert BE. High-dose, short-
duration ribavirin aerosol therapy in children
with suspected respiratory syncytial virus in-
fection. J Pediatr 1990; 117:313–20.
7. Cundy KC, Petty BG, Flaherty J, et al. Clinical
pharmacokinetics of cidofovir in human im-
munodeficiency virus–infected patients. Anti-
microb Agents Chemother 1995; 39:1247–52.
Reprints or correspondence: Dr. Albert Heim, Medizinische
Hochschule Hannover, Carl-Neuberg-Str. 1, D-30625 Hannover,
Germany (Heim.Albert@mh-hannover.de).
Clinical Infectious Diseases 2008; 47:731–2
� 2008 by the Infectious Diseases Society of America. All
rights reserved. 1058-4838/2008/4705-0026$15.00
DOI: 10.1086/590970
Glycerol and Dexamethasone
in Bacterial Meningitis in
Low-Income Countries:
Response to the Editorial
Commentary by Sa´ez-Llorens
and McCracken Jr.
To the Editor—Some of the comments
and observations by Sa´ez-Llorens and
McCracken Jr. [1] regarding the study by
Peltola et al. [2], in which the role of
glycerol and dexamethasone in acute bac-
terial meningitis has been evaluated, are
provocative and need discussion, partic-
ularly after the response by Peltola and
Roine [3].
Sa´ez-Llorens and McCracken Jr. [1]
raised ethical concerns about the use of
placebo and contend that, because of the
use of placebo, many children with bac-
terial meningitis were subjected to an un-
necessary risk of sequelae. Given the cur-
rent evidence, we do not believe that this
is true. A recent meta-analysis supports
the use of adjunctive corticosteroids for
children in high-income countries only
[4]. For children in low-income countries,
corticosteroids had no beneficial effect on
mortality (risk ratio [RR], 0.96; 95% CI,
0.78–1.18), severe hearing loss (RR, 1.04;
95% CI, 0.66–1.63), and short-term neu-
rological sequelae (RR, 1.08; 95% CI,
0.82–1.44). For children in low-income
countries, the use of corticosteroids was
not associated with either beneficial or
harmful effects [4].
In low-income countries, many of the
patients have underlying malnutrition and
frequent infections, and they often do not
approach a health care facility until late in
the course of an illness. These factors are
likely to affect cortisol production and lev-
els of cortisol in plasma and may affect
the outcome of a critical illness. Mal-
nourished children with or without acute
infection have hypercortisolemia and im-
paired clearance of exogenous cortisol [5–
7]. However, the expected metabolic ef-
fects of increased cortisol and other glu-
cocorticoid agonists are blunted in these
children, because of resistance to gluco-
corticoids secondary to downregulation of
glucocorticoid receptor protein, expres-
sion of an inactive form of the glucocor-
ticoid receptor protein, or repression of
phosphorylation of the glucocorticoid re-
ceptor/hormone complex [7, 8].
In a study from our center, serum cor-
tisol levels in patients who had acute bac-
terial meningitis were very high. Only 2
of 30 patients had serum cortisol levels in
the normal range (20 ng/mL in 1 patient
and 50 ng/mL in the other) [9]. The levels
correlated with the severity of illness as
assessed using the Glasgow Coma Scale.
We contend that the therapeutic use of
exogenous steroids in patients in low-in-
come countries who already have very
high endogenous corticosteroid secretion
may not be effective, and this may explain
the lack of benefit from dexamethasone

CORRESPONDENCE • CID 2008:47 (1 September) • 733
treatment that has been reported in studies
from developing countries [10]. Until
such time that new data or meta-analyses
are available to define the reasons for dif-
ferent outcomes in high-income versus
low-income countries and to identify
those children in low-income countries
who could benefit form corticosteroids,
the role of adjunctive corticosteroids for
the treatment of acute bacterial meningitis
in low-income countries remains un-
certain.
We have also conducted a study [11]
that used a protocol similar to that used
by Peltola et al. [2]. Our study was
approved by the ethics committee of our
institution without any concerns. We did
not find any difference in outcomes
among patients treated with glycerol, dex-
amethasone, or placebo. However, we
found that children who received glycerol
had a mean 3% elevation of serum os-
molality during the first 6 h after treat-
ment, and this elevation was sustained for
24 h. This could have contributed to low-
ered intracranial pressure and improved
cerebral perfusion and, therefore, reduc-
tion of sequelae associated with bacterial
meningitis [12].
A prerequisite for a “favorable” effect
of dexamethasone treatment in patients
who have bacterial meningitis is that it be
administered before the first dose of an-
tibiotic. In reality, a large number of pa-
tients in developing countries report to the
hospital late in the course of the disease
and may have received �1 antibiotics be-
fore reporting to a hospital [13]. There-
fore, not using dexamethasone as an ad-
junct is not unreasonable.
In the absence of clear proof of the ben-
efit of dexamethasone therapy for the
treatment of bacterial meningitis, the del-
eterious effect of a high cortisol level on
neurological outcome cannot be dis-
counted. Glucocorticoids produce a gen-
eralized metabolic vulnerability in neu-
rons that possess a high concentration of
corticosterone receptors [14]. The use of
dexamethasone in such a scenario could
perhaps raise ethical concerns.
Acknowledgments
Potential conflicts of interest. S.S. and P.S.:
no conflicts.
Sunit Singhi and Pratibha Singhi
Pediatric Emergency and Intensive Care Unit,
Advanced Pediatrics Centre, Postgraduate Institute
of Medical Education and Research,
Chandigarh, India
References
1. Sa´ez-Llorens X, McCracken GH Jr. Glycerol
and bacterial meningitis. Clin Infect Dis
2007; 45:1287–9.
2. Peltola H, Roine I, Fernandez J, et al. Adjuvant
glycerol and/or dexamethasone to improve the
outcome of childhood bacterial meningitis: a
prospective, randomized, double-blind, pla-
cebo-controlled trial. Clin Infect Dis 2007; 45:
1277–86.
3. Peltola H, Roine I. Glycerol and bacterial men-
ingitis: a response to the editorial commentary
by Sa´ez-Llorens and MC Cracken. Clin Infect
Dis 2008; 46:1123–4.
4. van de Beek D, de Gans J, McIntyre P, Prasad
K. Corticosteroids for acute bacterial menin-
gitis. Cochrane Database Syst Rev 2007; 1:
CD004405.
5. Laditan AA. Hormonal profiles in children
with progressively worsening nutritional
status. Hum Nutr Clin Nutr 1982; 36C:81–6.
6. Samuel AM, Kadival GV, Patel BD, Desai AG.
Adrenocorticosteroids and corticosteroid
binding globulins in protein calorie malnu-
trition. Am J Clin Nutr 1976; 29:889–94.
7. Manary MJ, Muglia LJ, Vogt SK, Yarasheski
KE. Cortisol and its action on the glucocor-
ticoid receptor in malnutrition and acute in-
fection. Metabolism 2006; 55:550–4.
8. Ray A, Prefontaine KE. Physical association
and functional antagonism between the p65
subunit of transcription factor NF-kappa B
and the glucocorticoid receptor. Proc Natl
Acad Sci U S A 1994; 91:752–6.
9. Singhi SC, Bansal A. Serum cortisol levels in
children with acute bacterial and aseptic men-
ingitis. Pediatr Crit Care Med 2006; 7:74–8.
10. Molyneux EM, Walsh AL, Forsyth H, Tembo
M, Mwenechanya J, Kayira K. Dexamethasone
treatment in childhood bacterial meningitis in
Malawi: a randomised controlled trial. Lancet
2002; 360:211–8.
11. Sankar J, Singhi S, Bansal A, Ray P, Singhi P.
Role of dexamethasone and oral glycerol in
reducing hearing and neurological sequelae in
children with bacterial meningitis. Indian Pe-
diatr 2007; 44:649–56.
12. Singhi S, Ja¨rvinen A, Peltola H. Increase in
serum osmolality is possible mechanism for
the beneficial effects of glycerol in childhood
bacterial meningitis. Pediatr Infect Dis (in
press).
13. Singhi S, Mohan Kumar D, Singhi PD, Sapru
S, Ganguli NK. Evaluation of polymerase
chain reaction (PCR) for diagnosis of Hae-
mophilus influenzae-B meningitis. Ann Trop
Paediatr 2002; 22:347–53.
14. Sapolsky RM. A mechanism for glucocorti-
coids toxicity in the hippocampus: increased
neuronal vulnerability to metabolic insults. J
Neurosci 1985; 5:1228–32.
Reprints or correspondence: Dr. Sunit Singhi, Dept. of Pedi-
atrics, Pediatric Emergency and Intensive Care, Advanced Pe-
diatrics Centre, Postgraduate Institute of Medical Education
and Research, Chandigarh 160012, India (dr_singhi
@yahoo.com).
Clinical Infectious Diseases 2008; 47:732–3
� 2008 by the Infectious Diseases Society of America. All
rights reserved. 1058-4838/2008/4705-0027$15.00
DOI: 10.1086/590971
Reply to Singhi et al.
To the Editor—In the concluding par-
agraph of our commentary regarding the
study by Peltola et al. [1], we stated “that
the most effective strategy of managing
bacterial meningitis and its associated se-
quelae is prevention by implementing
large-scale immunization strategies
against the common meningeal patho-
gens.…We urgently need a dedicated col-
laborative effort between vaccine manu-
facturers, philanthropic foundations,
global health organizations, and national
governments to make these vaccines avail-
able for those who need them most: the
infants and children living in less advan-
tageous areas of the world [2, pp.
1288–9].”
We do not believe that additional dis-
cussion of whether dexamethasone or
glycerol is more effective as adjunctive
therapy for bacterial meningitis in infants
and children from very low income
nations is a useful exercise. These agents
have a relatively small impact in those who
are most severely affected, compared with
the impact of prevention of disease by
vaccination.
Singhi and Singhi [3] took issue with
our concern regarding the use of placebo
in children with bacterial meningitis in the
study by Peltola et al. [1]. They misread
our commentary. We questioned the use

734 • CID 2008:47 (1 September) • CORRESPONDENCE
of placebo for the treatment of meningitis
in countries where disease due to Hae-
mophilus influenzae type b (Hib) predom-
inates. This was the case in the Latin
American countries included in the study
by Peltola et al. [1], in which 221 (45.7%)
of 484 patients had meningitis caused by
Hib. Stratifying the outcomes for patients
with disease due to Hib into dexametha-
sone versus non-dexamethasone recipi-
ents from the Peltola et al. [1] study il-
lustrates our point. Severe neurologic
sequelae were observed in 4 (4.5%) of 88
and 9 (9.5%) of 95 patients who received
dexamethasone or did not receive dexa-
methasone (difference is not statistically
significant), whereas profound hearing
loss was observed in 5 (5.5%) of 91 and
16 (17.8%) of 90 patients ( ), re-Pp .01
spectively. This was likely a concern for 2
institutions in Buenos Aires that decided
not to administer the placebo-placebo reg-
imen to their patients who were enrolled
in that study [1].
In a previous study, Singhi and Bansal
[4] showed that serum cortisol concen-
trations in children with bacterial men-
ingitis were high, especially in those with
very severe disease, as assessed using the
Glasgow Coma Scale. They contend that
the use of exogenous corticosteroids in
these patients would likely be ineffective,
but they provide no direct evidence to
support this contention. Their study [4]
used a protocol design similar to that of
Peltola et al. [1], and it was too under-
powered (56 patients divided among 4
therapeutic arms) to make any conclu-
sions. However, in the analysis of the data
from the Latin American study, Peltola
and colleagues demonstrated that dexa-
methasone or glycerol alone or the com-
bination of the 2 agents prevented hearing
impairment (measured at 80, 60, or 40
dB), regardless of the causative organism
or timing of adjunctive therapy (H. Pel-
tola, personal communication). The only
factor that independently predicted hear-
ing loss was the severity of illness at the
time of diagnosis; each lower score in the
Glasgow Coma Scale increased the risk of
impairment by 15%–20%.
As emphasized elsewhere [5], the only
effect of dexamethasone in the treatment
of meningitis is to modulate the secondary
meningeal inflammatory response that
follows the initial dose of a parenterally
administered, bactericidal antibiotic, and
the principal clinical benefit in children is
the reduction of hearing loss. Corticoste-
roids do not reverse CNS damage that has
already developed as a result of inflam-
mation, cerebral edema, increased intra-
cranial pressure, and vascular thrombosis.
This certainly applied to the children with
severe meningitis who were treated in Ma-
lawi [6], where undernutrition, concom-
itant HIV infection, young age, antimi-
crobials with poor activity, and delayed
presentation resulted in very high case-
fatality and long-term morbidity rates for
disease due to Hib and Streptococcus pneu-
moniae, regardless of whether dexametha-
sone was administered.
Acknowledgments
Potential conflicts of interest. X.S.-L. and
G.H.M.: no conflicts.
Xavier Sa´ez-Llorens1
and George H. McCracken, Jr.2
1University of Panama School of Medicine,
Hospital del Nin˜o, Panama City, Panama;
and 2University of Texas Southwestern
Medical Center at Dallas, Dallas
References
1. Peltola H, Roine I, Fernandez J, et al. Adjuvant
glycerol and/or dexamethasone to improve the
outcome of childhood bacterial meningitis: a
prospective, randomized, double-blind, pla-
cebo-controlled trial. Clin Infect Dis 2007; 45:
1277–86.
2. Sa´ez-Llorens X, McCracken GH. Glycerol and
bacterial meningitis. Clin Infect Dis 2007; 45:
1287–9.
3. Singhi S, Singhi P. Glycerol and dexamethasone
in bacterial meningitis in low-income coun-
tries: response to the editorial commentary by
Sa´ez-Llorens and McCracken Jr. Clin Infect Dis
2008; 47:732–3 (in this issue).
4. Singhi SC, Bansal A. Serum cortisol levels in
children with acute bacterial and aseptic men-
ingitis. Pediatr Crit Care Med 2006; 7:74–8.
5. McCracken G. Rich nations, poor nations and
bacterial meningitis. Lancet 2002; 360:183.
6. Molyneux EM, Walsh AL, Forsyth H, Tembo
M, Mwenechanya J, Kayira K. Dexamethasone
treatment in childhood bacterial meningitis in
Malawi: a randomized controlled trial. Lancet
2002; 360:211–8.
Reprints or correspondence: Dr. George H. McCracken, Jr.,
University of Texas Southwestern Medical Center at Dallas,
5323 Harry Hines Blvd., F3-202, Dallas, TX 75390-9063
(George.mccracken@utsouthwestern.edu).
Clinical Infectious Diseases 2008; 47:733–4
� 2008 by the Infectious Diseases Society of America. All
rights reserved. 1058-4838/2008/4705-0028$15.00
DOI: 10.1086/590973