Ribavirin for the Treatment of HCPS • CID 2004:39 (1 November) • 1307
M A J O R A R T I C L E
Placebo-Controlled, Double-Blind Trial
of Intravenous Ribavirin for the Treatment
of Hantavirus Cardiopulmonary Syndrome
in North America
Gregory J. Mertz,1Lil Miedzinski,6Diane Goade,1Andrew T. Pavia,2Brian Hjelle,1Christine O. Hansbarger,1
Howard Levy,1,aFrederick T. Koster,1Kenneth Baum,3Adeline Lindemulder,6Wenquan Wang,4Laura Riser,4
Humberto Fernandez,5and Richard J. Whitley,4for the Collaborative Antiviral Study Group
1University of New Mexico, Albuquerque;
at Birmingham;5ICN Pharmaceuticals, Costa Mesa, California; and
2University of Utah, Salt Lake City;
3University of Colorado, Denver;
6University of Alberta, Edmonton, Alberta, Canada
4University of Alabama
intravenous ribavirin for the treatment of hantavirus cardiopulmonary syndrome (HCPS) were inconclusive.
Subjects with suspected HCPS in the prodrome or cardiopulmonary phase but without shock were
eligible for randomization to receive either intravenous ribavirin (33 mg/kg [?2 g], followed by 16 mg/kg [?1
g] given every 6 h for 4 days and by 8 mg/kg [?.5 g] given every 8 h for 3 days) or placebo (administered for
7 days or until the initial Sin Nombre virus antibody test result was confirmed to be negative). The primary
outcome was survival at day 28 of the study without the need for extracorporeal membrane oxygenation(ECMO).
Thirty-six subjects were enrolled in the trial from March 1996 through July 2001, at which point the
study was terminated prematurely because of both the slow rate of accrual of subjects and the findings of a futility
analysis. Of the 36 subjects enrolled, 23 (all of whom were enrolled during the cardiopulmonary stage of HCPS)
had HCPS confirmed by serologic testing. The severity of illness at entry into the study was similar among the
10 subjects with HCPS who received ribavirin and the 13 subjects with HCPS who received placebo. Theproportion
of subjects who survived and who did not require ECMO was similar among ribavirin recipients and placebo
recipients (70% vs. 62%, respectively); 2 ribavirin recipients and 2 placebo recipients died, including 3 of 7 subjects
treated with ECMO. The frequency of adverse events, including anemia, was similar between treatment groups.
The rate of accrual of subjects in the present study was inadequate to clearly assess the safety
or efficacy of ribavirin in the treatment of HCPS. However, ribavirin was well tolerated, and the lack of trends
supporting the use of intravenous ribavirin suggests that it is probably ineffective in the treatment of HCPS in
the cardiopulmonary stage.
Ribavirin is active in vitro against hantaviruses, but the findings of an open trial of the use of
Since the recognition of hantavirus cardiopulmonary
syndrome (HCPS) in 1993, HCPS has been identified
throughout most of North and South America and in
Panama. Known also as “hantavirus pulmonary syn-
Received 23 April 2004; accepted 16 June 2004; electronically published 11
aPresent affiliation: Now also affiliated with Acute Care–US Medical Division,
Eli Lilly, Indianapolis, Indiana.
Investigators and institutions who participated in the study are listed at the
end of the text.
Reprints or correspondence: Dr. Gregory J. Mertz, Internal Medicine, MSC10
5550, 1 University of New Mexico, Albuquerque, New Mexico 87131-0001
Clinical Infectious Diseases2004;39:1307–13
? 2004 by the Infectious Diseases Society of America. All rights reserved.
drome” (HPS), we prefer the term “hantavirus cardio-
pulmonary syndrome” because almost all fatalities as-
sociated with this syndrome are caused by cardiogenic
shock [1, 2]. In the United States, 379 cases of HCPS
have been reported through 1 September 2004, for a
case-fatality rate of 36% (http://www.cdc.gov/ncidod/
Canada, 47 cases of HCPS were reported through 31
December 2003, for a case-fatality rate of 36% (H. Art-
sorb, unpublished data). Sin Nombre virus (SNV) is
the etiologic agent of all cases of HCPS in Canada and
of all but a handful of cases of HCPS in the United
States, including all of the cases that occurred among
subjects enrolled in the present study.
Ribavirin is active in vitro against all hantaviruses
by guest on October 18, 2015
1308 • CID 2004:39 (1 November) • Mertz et al.
tested [3, 4]. Intravenous ribavirin has been reported to be
effective in the treatment of hemorrhagic fever with renal syn-
drome (HFRS), which is caused by the prototype hantavirus
Hantaan virus . In an open-label trial of intravenous riba-
virin for the treatment of HCPS, which was conducted in the
were used as control subjects . Treatment was often delayed
by the need to ship the study drug from regional distribution
centers, and treatment was not allowed during the prodrome
phase of HCPS, in part because rapid serologic testing was not
available. Rapid serologic testing subsequentlybecameavailable
and wasshownto behighlysensitiveduringboththeprodromal
and cardiopulmonary phases of HCPS . The current trial
was designed to evaluate the efficacy and safety of ribavirin in
the treatment of HCPS and to allow enrollment of subjects
with cases of HCPS that were in either the prodromal phase
or the cardiopulmonary phase.
SUBJECTS AND METHODS
?12 years of age and who had suspected or serologically con-
firmed acute hantavirus disease that was in the prodromal or
cardiopulmonary phase were eligible for enrollment in the
study. A presumptive diagnosis of HCPS required the presence
of a compatible prodromal illness with fever of !12 days’ du-
ration; myalgia, with or without nausea and vomiting, abdom-
inal pain, or diarrhea; absence of upper respiratory tract symp-
toms at the onset of symptoms; a platelet count of !150,000
cells/mm3; and, for patients with HCPS in the cardiopulmonary
phase, immunoblasts comprising at least 10% of peripheral
blood lymphocytes. The study drug was not administered to
premenopausal women until the result of a serum or urine
pregnancy test was confirmed to be negative.
Criteria for exclusion from the study included pregnancy or
breast-feeding, a likely diagnosis other than HCPS, immuno-
compromised status, receipt of systemic corticosteroids within
30 days before enrollment in the study, a mean arterialpressure
(MAP) of !60 mm Hg for 2 h despite optimal medical man-
agement, a cardiac index of !2.1 L/min/m2, arterial oxygen
pressure of !65 mm Hg in intubated subjects receiving 100%
oxygen, or the presence of unilateral pulmonary infiltrates that
did not become bilateral within 24 h. Subjects providedwritten,
informed consent, which included an agreement to practice
birth control for 6 months. The clinical trial was approved by
the institutional review boards of all participating institutions.
Subjects received either intravenous ri-
bavirin (33 mg/kg [?2 g] as a loading dose, followed by 16
mg/kg [?1 g] given q6h for 4 days and by 8 mg/kg [?500
mg] given q8h for 3 days) or placebo for 7 days. This regimen
was found to be effective in the treatment of HFRS . Ran-
domization to treatment was done in groups of 4, with 2 sub-
Males and nonpregnant females who were
jects randomized to receive ribavirin and the other 2 subjects
randomized to receive placebo. The study drug was kept in the
research pharmacy at each site, and identification numbers (in
sequence) were assigned to subjects at the time of their en-
rollment in the trial. All staff at each study site were blinded
as to treatment assignment. The study drug was discontinued
if the result of testing for SNV antibody was negative or when
extracorporeal membrane oxygenation (ECMO) was initiated.
Ribavirin and matching placebo were provided by ICN Phar-
maceutical (Costa Mesa, CA)
Serologic testing for hantavirus.
by use of a strip immunoblot assay. Testing was performed at
the University of New Mexico in Albuquerque .
Primary and secondary end points.
of the study was survival at day 28 after study entry. When
centers participating in the study began using ECMO as salvage
therapy for patients for whom death appeared to be imminent,
the primary end point was revised to survival at day 28 after
study entry without the use of ECMO. Secondary efficacy eval-
uations included determination of the Murray lung score (a
lung-injury score based on hypoxemia, compliance, positive
end-expiratory pressure, and the number of quadrantsshowing
alveolar consolidation on chest radiographs); development,du-
ration, and severity of shock; duration of hospitalization and
duration of stay in the intensive care unit; and the peak level
of lactic acid and the duration of this peak level. The safety
parameters that were evaluated encompassed the development,
duration, and severity of anemia; elevation of transaminaseand
creatinine levels; and the development of pancreatitis and el-
evations in the serum level of amylase.
Serial clinical and laboratory evaluations.
followed daily on days 1–7 of the study, on days 14, 28, and
84, and at 6 months, by physical examination and assessment
of vital signs. On each of the aforementioned days, except for
day 6, a complete blood cell count with differential was deter-
mined; electrolyte, blood urea nitrogen,creatinine,alanineami-
notransferase, aspartate aminotransferase, bilirubin, albumin,
amylase, and uric acid levels, and serum levels of lactate were
measured; and a chest radiograph was obtained. A pregnancy
test was performed on day 84 and again at 6 months. Spirom-
etry was performed and diffusion capacity of the lung for car-
bon monoxide was measured on day 7, as well as at subsequent
visits, if the patient did not require mechanical ventilation.
Urinalysis and a coagulation screening test were performed at
study entry and on days 3, 5, 7, 28, and 84 of the study.
Data safety and monitoring board (DSMB).
established at the National Institutes of Health (Bethesda,MD).
After the enrollment of 15 patients with confirmed infection,
a review of the safety and efficacy data was performed in which
members of the DSMB were blinded to the study treatment
Serum samples weretested
A DSMB was
by guest on October 18, 2015
Ribavirin for the Treatment of HCPS • CID 2004:39 (1 November) • 1309
firmed hantaviral infection.
Demographic characteristics, at baseline, of patients with con-
(n p 10)
(n p 13)P
Age, median years (range)
Race or ethnicity
Days from onset of symptoms,a
fraction, where room air is 0.21 and 100% oxygen is 1.0.
aAt randomization to treatment.
bData are for 8 patients. The 2 remaining patients were breathing room air at
cData are for 6 patients. The 7 remaining patients were breathing room air at
Data are no. (%) of patients, unless indicated otherwise. Fio2, inspired oxygen
therapy for hantavirus cardiopulmonary syndrome. Solid line, ribavirin re-
cipients (); broken line, placebo recipients (n p 10
).n p 13
assigned; this review did not result in an alteration of the study
design or in premature termination of the trial because of drug
When the study was designed, we as-
sumed that the mortality rate among subjects with hantavirus
infection who were given placebo would be ?50%. Sample-
size calculations were performed with 80% power and a sig-
nificance level of 5% (by a 2-sided test), to detect an overall
reduction of 50% in the mortality rate among subjects treated
with ribavirin, compared with subjects given placebo. A total
of 130 subjects (65 in each group) was required.
Characteristics at baseline were recorded for all subjects. For
the primary and secondary efficacy end points, descriptive sta-
tistics were collected only for those patients with confirmed
SNV infection. For safety end points, descriptive statistics were
collected for all subjects. For continuous variables, the Wil-
coxon rank sum test was applied, and, for categorical data,
Fisher’s exact test was applied to test the difference between
the 2 treatment groups. The time-to-event distribution was
estimated using the product-limit method of Kaplan and Meier
and was compared using the log-rank test .
From March 1996 through July 2001, 36 subjects were enrolled
in the present study, including 26 subjects at the University of
New Mexico, 7 subjectsattheUniversityofAlberta(Edmonton,
Alberta, Canada), 2 subjects at the University of Utah (Salt
Lake City), and 1 subject at the University of Colorado
(Denver). Ten of the 23 subjects with confirmed SNV infection
received intravenous ribavirin, and 13 received placebo. The
demographic characteristics of the ribavirin recipients and the
placebo recipients were similar at baseline (table 1). Thirteen
subjects had negative results of serologic tests for SNV, and
administration of the study drug was discontinued after a me-
dian of 3 doses (range, 1–11 doses) had been given, when the
results of serologic testing became available. Because accrual of
subjects occurred at a slow rate, thestudywasterminatedbefore
the target number of enrolled subjects was reached.
Survival for 28 days without ECMO.
The proportion of
by guest on October 18, 2015
1310 • CID 2004:39 (1 November) • Mertz et al.
Outcomes for subjects with confirmed hantaviral
(n p 10)
(n p 13)P
Survival at day 28 of the study,
no. (%) of patients
Time to either death or ECMO
initiation, median h (range)
4 (3–15) 24 (4–64).85b
aBy Fisher’s exact test.
bBy log-rank test.
ECMO, extracorporeal membrane oxygenation.
denotes the mean) in a trial of intravenous ribavirin therapy for hantavirus cardiopulmonary syndrome.
Mean arterial pressure of ribavirin recipients (squares; the solid line denotes the mean) and placebo recipients (triangles; the broken line
patients who survived without the need for ECMO was similar
among ribavirin recipients (7 [70%] of 10 subjects)andplacebo
recipients (8 [62%] of 13 subjects) (figure 1). The median
duration from the time of administration of the first dose of
the study drug to either death or initiation of ECMO was 4 h
(range, 3–15 h) for ribavirin recipients versus 24 h (range, 4–
64 h) for placebo recipients (P p .85
bavirin recipients and 2 placebo recipients died, including 3 of
7 subjects who were treated with ECMO (table 2). Overall, 8
(80%) of 10 ribavirin recipients and 11 (85%) of 13 placebo
recipients survived at day 28 of the study. No deaths occurred
after day 28 of the study.
Secondary efficacy outcomes.
differences or trends in any of the secondary outcome mea-
; log-rank test). Two ri-
There were no significant
surements. The MAPs (figure 2) and the Murray lung scores
(figure 3) were similar in the 2 treatment groups. Lacticacidosis
(lactic acid level, 12 mmol/L) developed in 6 (60%) of 10
ribavirin recipients and in 9 (69%) of 13 placebo recipients
( ). There were no significant differences in either the
P p .69
median duration of lactic acidosis (median duration in each
treatment group, 2 days;
P p .81
lactic acid among ribavirin recipients versus placebo recipients
(4.4 vs. 2.8 mmol/L, respectively;
Treatment with ECMO.
Seven subjects, including 3 riba-
virin recipients and 4 placebo recipients, were treated with
ECMO. Six of these 7 subjects received care at the University
of New Mexico, and 1 subject received care at the University
of Alberta. Table 3 shows the hemodynamic parameters and
laboratory values that documented the need for salvagetherapy
and that were obtained just before initiation of ECMO.
of adverse events between treatment groups. There was a trend
toward a higher incidence of anemia among ribavirin recipients,
but this trend was limited to cases of mild to moderate anemia
(hemoglobin level, 7.0–9.4 g/dL). Mild to severe anemia (he-
moglobin level, 6.5–9.4 g/dL) was recorded for 4 (40%) of 10
ribavirin recipients and for 4 (31%) of 13 placebo recipients;the
single case of severe anemia (hemoglobin level, 6.5–6.9 g/dL)
occurred in a placebo recipient. There was no evidence of hep-
atotoxicity, nephrotoxicity, or development of pancreatitis as a
result of treatment with ribavirin. With regard to increases in
serum levels of amylase, severe toxicity (2.1–5 times the upper
) or the median peak level of
P p .60
by guest on October 18, 2015
Ribavirin for the Treatment of HCPS • CID 2004:39 (1 November) • 1311
for hantavirus cardiopulmonary syndrome.
Median Murray lung score of ribavirin recipients (solid bars)
umenting the need for salvage therapy before initiation of extra-
corporeal membrane oxygenation (ECMO).
Hemodynamic parameters and laboratory values doc-
ECMO. MAP , mean arterial pressure.
aMedian value, 159 beats/min.
bMedian value, 2.1 L/min/m2.
cMedian value, 68 mm Hg.
dMedian value, 5.2 mmol/L.
eMedian value for the ratio of PaO2to Fio2, 77. PaO2is the arterial oxygen
pressure in millimeters of mercury. Fio2is the inspired oxygen fraction, where
room air is 0.21 and 100% oxygen is 1.0.
Values in bold are life-threatening values mandating institution of
limit of the rangeconsideredtobenormal)developedin1(10%)
of 10 ribavirin recipients, whereas mild toxicity (1.6–2 times the
upper limit of the range considered to be normal) developed in
1 (8%) of 13 placebo recipients and severe toxicity developed in
3 (23%) of 13 placebo recipients.
The present trial is, to the best of our knowledge, the first
placebo-controlled, double-blind trial of intravenous ribavirin
or trends in overall survival, the proportion of subjects who
survived at day 28 of the study without ECMO, or the time
from administration of the first dose of the study drug to either
death or initiation of ECMO. Between study groups, there were
no significant differences or trends regarding measuresofshock
(including MAP and presence of lactic acidosis) andrespiratory
failure (including Murray lung scores).
Because Huggins et al.  reported clinical benefit resulting
from the use of intravenous ribavirin for treatment of HFRS
in a placebo-controlled trial, one might speculate as to why we
were not even able to demonstrate trends toward the efficacy
of intravenous ribavirin therapy for HCPS. First, we do not
think that it was because of any difference in the ribavirin
susceptibility of SNV, compared with that of Hantaan virus
(the primary cause of HFRS), because the ID50of ribavirin is
similar for Sin Nombre and Hantaan viruses [3, 4] (R. Medina
and B.H., unpublished data).
A second possibility is that we missed a true difference in
efficacy because of the small number of subjects enrolled in
the present trial. Although the study did not have adequate
power to determine efficacy, a futility analysis predicted that
we would have needed 1500 patients per group to show any
differences as predicated on the trends among the 23 confirmed
subjects. The lack of trends or significant differences in overall
survival, survival without ECMO, or any of the secondary ef-
ficacy parameters suggests that ribavirin probably would not
have been found to produce a large treatment effect, even if
we had reached our goals regarding accrual of subjects.
A third possibility for an apparent difference in the efficacy
of intravenous ribavirin therapy for HFRS and HCPS is asso-
ciated with the rate of disease progression. In both syndromes,
a febrile prodrome phase precedes the phase in which end-
organ failure occurs and in which there is risk of shock and
death. However, once the second phase begins, the rate of pro-
gression of the disease and time to death are significantly
shorter for individuals with HCPS than for individuals with
HFRS . In the study of HFRS, the median time from ini-
tiation of the study drug to death was 7 days (range, 2–10 days)
for placebo recipientsand 5days(range,5–6days)forrecipients
of intravenous ribavirin . In contrast, in the present trial,
the median time from administration of the first dose of the
study drug to death or initiation of ECMO was only 4 h (range,
3–15 h) for subjects given ribavirin and 24 h (range, 4–64 h)
for subjects given placebo.
When we designed the present trial, we envisioned enrolling
most subjects during the prodrome phase of HCPS, to allow
for administration of the study drug for several days before
development of progression to the cardiopulmonary phase of
the syndrome and death. Unfortunately, despite the availability
of rapid serologic testing and the high sensitivity of serologic
testing during the prodrome phase [1, 6, 8], we were unable
to enroll any subjects during the prodrome phase of HCPS.
Although many of the study subjects sought medical care dur-
ing the prodrome phase, HCPS generally was not considered
as a diagnosis at this time, and patients were not referred until
after the onset of the cardiopulmonary phase of the syndrome.
Once the cardiopulmonary phase of HCPS begins, we believe
by guest on October 18, 2015
1312 • CID 2004:39 (1 November) • Mertz et al.
that the rate of disease progression and the time to death are
too rapid for ribavirin to be of benefit.
Ribavirin treatment is associated with potentially serious ad-
verse effects, including anemia and the potential for teratoge-
nicity if used in pregnant women. Although the number of
subjects enrolled was too small to conclude that ribavirin is
safe, we were reassured by the low rate of adverse effects noted
in the present study. There were no significant differences in
the frequency of adverse events, including anemia, between the
2 treatment groups. We were able to obtain the results of se-
rologic testing within 24 h after testing at the University of
New Mexico and within 48–72 h after testing at the other sites,
so exposure to ribavirin was limited to a median of 3 doses
(range, 1–11 doses) for persons without HCPS. As such, among
seronegative subjects, the risk of adverse effects was probably
minimized by our ability to discontinue the study drug after a
In the present open-labeltrial,amongthe128subjectstreated
with ribavirin, pancreatitis developed in 5 subjects and possible
pancreatitis developed in 2 subjects for whom safety data were
available . It was reassuring that there was no difference in
the incidence of elevated serum levels of amylase between ri-
bavirin recipients and placebo recipients in the present trial.
Although the number of subjects enrolled in our study was
small, the data from the present study do not support a sig-
nificant role for ribavirin in the development of pancreatitis.
The present trial was designed in 1994, at a time when we
had only slightly more than 1 year of clinical experience with
HCPS, before ECMO was used as rescue therapy, and before
anyone had experience in conducting a controlled trial of treat-
ment for HCPS. One of the primary benefits of the present
trial is what it has taught us about the design and conduct of
controlled trials of treatment for HCPS.
First, our efforts to enroll subjects during the prodrome
bocytopenia, but without waiting for the results of IgG and IgM
antibody testing—were not successful. All 12 subjects who met
the study-entry criteria for suspected hantavirusintheprodrome
phase subsequently were found to beseronegativeforhantavirus.
This rate of enrollmentofsubjectswithoutHCPSisunacceptably
high, particularly for a drug with an adverse event profile such
as that for ribavirin. Therefore, future studies should exclude
subjects with suspected HCPS in the prodrome phase who lack
confirmation of serologic test results, unless the study drug is
considered to have minimal associated risk and/or we can im-
prove the specificity of the clinical diagnosis.
Second, we learned that we could identify subjects with
HCPS in the cardiopulmonary phase on the basis of clinical
and laboratory criteria but,ofimportance,beforeserologiccon-
firmation. This was a matter of concern at the start of the study,
because, in the open-label study of persons with presumed
HCPS in the cardiopulmonary phase, infection was serologi-
cally confirmed for only 30 (21%) of 140 subjects who were
treated with ribavirin . Of the remaining 110 subjects, 105
were seronegative and 5 did not have serologic testing per-
formed. In the present study, 23 (96%) of 24 subjects who were
enrolled in the trial during the cardiopulmonary phase were
found to be seropositive.
Third, the rate of subject accrual in the present study was
slow, even in epidemic years. The main reasons for the slow
accrual of subjects were (1) that patients who presented with
shock were almost always excluded from the study on the basis
of our exclusion criteria and a belief that the need for salvage
therapy was imminent, and (2) that the treating physician
thought that ribavirin was unlikely to work quickly enough to
benefit a patient with HCPS who was already in shock. Future
trials involving subjects with HCPS in the cardiopulmonary
phase should allow study enrollment of subjects who are in
shock and should evaluate agents with a rapid onset of action.
On the basis of current knowledge, there are several prom-
ising treatments that could be evaluated in controlled trials.
First, patients with serious or fatal HCPS in the United States
and Chile have significantly lower titers of serum neutralizing
antibody on the day of admission to the hospital, compared
with patients who have diseases of mild severity  (P. Vial,
unpublished data), and neutralizing antibody is protective in
an animal model of lethal Andes virus infection . Therefore,
there is a rationale for evaluation of administration of neu-
tralizing antibody for the treatment of HCPS. There is also
strong evidence that suggests an immunologic basis for the
cardiopulmonary phase,and thereisanongoingcontrolled
trial in Chile that is evaluating the use of corticosteroids for
patients who present in the cardiopulmonary phase.
In summary, we did not meet our goal for enrollment of
subjects, and the present study lacked adequate power to de-
termine whether ribavirin was safe and effective in the treat-
ment of HCPS. However, the lack of trends favoring use of
ribavirin and the rapid progression to death or initiation of
ECMO after initiation of the study drug suggest that, although
intravenous ribavirin was well tolerated, it probably is not ef-
fective for the treatment of HCPS after the onset of the car-
diopulmonary phase. We were unable to enroll subjects during
the prodrome phase, and efforts to do so that were based solely
on clinical presentation and presence of thrombocytopenia led
to enrollment of subjects without SNV infection. With only a
small associated risk of enrollment of subjects without han-
tavirus infection in the study, we were able to identify subjects
with disease in the cardiopulmonary phase; however, accrual
was slowed by the exclusion of subjects with shock.
by guest on October 18, 2015
Ribavirin for the Treatment of HCPS • CID 2004:39 (1 November) • 1313 Download full-text
MEMBERS OF THE COLLABORATIVE
ANTIVIRAL STUDY GROUP RIBAVIRIN
CONTROLLED TRIAL FOR HCPS
Participating centers (locations) and members of the Collab-
orative Antiviral Study Group are as follows: University of New
Mexico (Albuquerque, NM): G.J.M., D.G., B.H., C.O.H., H.L.,
F.T.K., Suzanne Popejoy, Joseph Hubbard, and Karl Johnson;
University of Alberta (Edmonton, Alberta, Canada): L.M. and
A.L.; University of Utah (Salt Lake City): A.T.P. and John C.
Christenson; University of Alabama at Birmingham: R.J.W.,
L.R., Lynette Sherrill, W.W., and Mark Carpenter; ICN Phar-
maceuticals (Costa Mesa, CA): H.F.; University of Colorado
(Denver): K.B.; Gallup Indian Medical Center (Gallup, NM):
Bruce Tempest; Johns Hopkins University (Baltimore, MD):
Ray Reid; Division of Microbiology and Infectious Diseases,
National Institute of Allergy and Infectious Diseases, National
Institutes of Health (Bethesda, MD): Catherine Laughlin,Walla
Dempsey, and Thelma Gaither.
We thank Ms. Cindy Wootton, for help in preparation of themanuscript;
Suzanne Popejoy, for assistancewithextracorporealmembraneoxygenation
data; and Dr. Jonathan Iralu and the physicians working with participating
centers, for referring patients with suspected hantavirus cardiopulmonary
National Institute of Allergy and Infectious Dis-
eases, National Institutes of Health (contract NO-1-AI-30025; Bethesda,
Potential conflict of interest.
H.F. is an employee of ICN Pharmaceu-
ticals (Costa Mesa, CA). All other authors: No conflict.
1. Mertz GJ. Bunyaviridae: bunyaviruses, phleboviruses, nairovirusesand
hantaviruses. In: Richmann DD, Whitley RJ, Hayden FG, eds. Clinical
virology. 2nd ed. Washington, DC: American Society for Microbiology
2. Hallin GW, Simpson SQ, Crowell RE, et al. Cardiopulmonary mani-
festations of the hantavirus pulmonary syndrome. Crit Care Med
3. Huggins JW, Hsiang CM, Cosgriff TM, et al. Prospective,double-blind,
concurrent, placebo-controlled clinical trial of intravenous ribavirin
therapy for hemorrhagic fever with renal syndrome. J Infect Dis
4. Johnson KM. Ribavirin treatment of arenavirus, hantavirus, pneumo-
virus and paramyxovirus disease; topical and systemictherapy.In:Mills
J, Corey L eds. Antiviral chemotherapy: new directions for clinical
application and research. Vol 3. Englewood Cliffs, NJ: PTR Prentice
5. Chapman LE, Mertz GJ, Peters CJ, et al. Intravenous ribavirin for
hantavirus pulmonary syndrome: safety and tolerance during 1 year
of open-label experience. Antivir Ther 1999;4:211–9.
6. Hjelle B, Glass GE. Outbreak of hantavirus infection in the Four Cor-
ners region of the United States in the wake of the 1997–1998 El
Nin ˜o–southern oscillation. J Infect Dis 2000;181:1569–73.
7. Lee ET. Statistical methods or survival data analysis. 2nd ed. New York:
John Wiley and Sons, 1992.
8. Bharadwaj M, Nofchissey R, Goade D, Koster F, Hjelle B. Humoral
immune responses in the hantavirus cardiopulmonary syndrome. J
Infect Dis 2000;182:43–8.
9. Custer DM, Thompson E, Schmaljohn CS, Ksiazek TG, Hooper JW.
Active and passive vaccination againsthantaviruspulmonarysyndrome
with Andes virus M genome segment-based DNA vaccine. J Virol
10. Mertz GJ, Hjelle BL, Williams TM, Koster FT. Host responses in the
hantavirus cardiopulmonary syndrome. In: Saluzzo JF, Dodet B, eds.
Factors in the emergence and control of rodent-borne viral diseases.
Paris: Elsevier, 1999;133–7.
by guest on October 18, 2015