The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne

Correspondence 1. Cassileth PA, Harrington DP, Appelbaum FR, et al. Chemotherapy

compared with autologous or allogeneic bone marrow transplantation in
the management of acute myeloid leukemia in first remission. N Engl J
Med 1998;339:1649-56.

To the Editor: In the study by Cassileth et al., the mor-

Treatment of Acute Myeloid Leukemia
To the Editor: Cassileth et al. (Dec. 3 issue)1 report the
results of a large study comparing chemotherapy with au-
tologous or allogeneic bone marrow transplantation in the
management of acute myeloid leukemia in first remission.
Their study showed that autologous or allogeneic bone
marrow transplantation offered no advantage over inten-
sive chemotherapy in terms of overall survival. However,
the study included patients with a variety of risk factors.
Although there is no established prognostic index for
acute myeloid leukemia, we believe that most institutions
do not perform bone marrow transplantation in patients
with either the M2 or the M3 type of acute myeloid leu-
kemia in first remission, according to the French–Ameri-
can–British (FAB) classification, unless there are unfavor-
able factors, such as a high initial white-cell count or
difficulty in achieving remission.
What we want to know is, who really requires bone mar-
row transplantation in first remission? In other words, for
which patients should we reserve bone marrow transplan-
tation until they have a relapse? Characterizing the pa-
tients who have had relapses with chemotherapy and in
whom bone marrow transplantation as second-line thera-
py has failed will be informative in identifying eligible pa-
tients for further randomized studies. In such patients,
bone marrow transplantation as consolidation therapy may
be of benefit.
tality rate at 100 days in the autologous-transplantation
group was 14 percent; most of these deaths were due to
infection or graft failure. Autologous marrow grafts were
purged with perfosfamide. This technique has been asso-
ciated with delayed engraftment in a dose-dependent man-
ner, even in patients with neoplasms that originated out-
side the bone marrow.1 In the study by Cassileth et al.,
purging with perfosfamide may have contributed to early
death, and thus to the lack of a benefit in terms of leuke-
mia-free survival and the relapse rate. Their results should
be compared with those of studies in which unpurged
marrow was used.2-4 The conclusions of Cassileth et al.
should be viewed with caution when applied to the trans-
plantation of unpurged cellular products.
JAVIER PÉREZ-CALVO, M.D.
ANTONIO BRUGAROLAS, M.D.
University Clinic of Navarre
31008 Pamplona, Spain
1. Shpall EJ, Jones RB, Blast RC Jr, et al. 4-Hydroperoxycyclophospha-
mide purging of breast cancer from the mononuclear cell fraction of bone
marrow in patients receiving high-dose chemotherapy and autologous mar-
row support: a phase I trial. J Clin Oncol 1991;9:85-93.
2. Zittoun RA, Mandelli F, Willemze R, et al. Autologous or allogeneic
bone marrow transplantation compared with intensive chemotherapy in
acute myelogenous leukemia. N Engl J Med 1995;332:217-23.
3. Harousseau JL, Cahn JY, Pignon B, et al. Comparison of autologous
bone marrow transplantation and intensive chemotherapy as postremission
therapy in adult acute myeloid leukemia. Blood 1997;90:2978-86.
4. Burnett AK, Goldstone AH, Stevens RM, et al. Randomised compari-
son of addition of autologous bone-marrow transplantation to intensive
chemotherapy for acute myeloid leukaemia in first remission: results of
MRC AML 10 trial. Lancet 1998;351:700-8.

YOSHINOBU KANDA, M.D. To the Editor: In the study by Cassileth et al., patients
AKIYOSHI MIWA, M.D. were randomly assigned to undergo bone marrow trans-
ATSUSHI TOGAWA, M.D. plantation or to receive high-dose cytarabine chemothera-
International Medical Center of Japan py after one course of consolidation therapy consisting of
Tokyo 162-8655, Japan two days of idarubicin and five days of cytarabine. No sig-

INSTRUCTIONS FOR LETTERS TO THE EDITOR

Letters to the Editor are considered for publication (subject to editing and abridgment) provided they do not contain material that has been
submitted or published elsewhere. Please note the following: •Your letter must be typewritten and triple-spaced. •Its text, not including
references, must not exceed 400 words (please include a word count). •It must have no more than five references and one figure or table.
•It should not be signed by more than three authors. •Letters referring to a recent Journal article must be received within four weeks of its
publication. •Please include your full address, telephone number, and fax number (if you have one). •You may send us your letter by post, fax,
or electronic mail.
Our address: Letters to the Editor • New England Journal of Medicine • 10 Shattuck St. • Boston, MA 02115
Our fax numbers: 617-739-9864 and 617-734-4457
Our e-mail address: letters@nejm.org
We cannot acknowledge receipt of your letter, but we will notify you when we have made a decision about publication. We are unable to
provide prepublication proofs. Please enclose a stamped, self-addressed envelope if you want unpublished material returned to you.
Financial associations or other possible conflicts of interest must be disclosed. Submission of a letter constitutes permission for the Massachu-
setts Medical Society, its licensees, and its assignees to use it in the Journal’ s various editions (print, data base, and optical disk) and in
anthologies, revisions, and any other form or medium.

1436 · May 6 , 19 9 9

The New England Journal of Medicine

Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
 C ORR ES POND ENCE
nificant advantage in terms of disease-free survival was
shown for bone marrow transplantation. In fact, the over-
all survival rate was better for the patients who received
high-dose cytarabine chemotherapy than for those who
underwent autologous marrow transplantation.
Although bone marrow transplantation has an aspect of
immunotherapy through its graft-versus-leukemia effect,
transplantation itself, especially autologous transplantation,
is not a cytocidal therapy but a means of hematopoietic res-
cue after myeloablative chemotherapy. Therefore, the thera-
peutic effect of transplantation should be considered in the
context of the remission-induction and consolidation che-
motherapy and the preparative regimens the patients receive.
In most previous prospective studies, bone marrow
transplantation was performed after no consolidation ther-
apy, or after only one cycle of it, and these studies showed
no or only marginal benefits of transplantation. However,
two large-scale, randomized trials of autologous marrow
transplantation for acute myeloid leukemia, in which the
transplant was given after at least one cycle of intensive
consolidation chemotherapy, showed a significant improve-
ment in disease-free survival in the patients assigned to re-
ceive transplants.1,2 Although these studies failed to show
any difference in overall survival, a study in which patients
undergo transplantation after multiple cycles of intensive
consolidation chemotherapy might well demonstrate an
advantage of transplantation. The effect of transplantation
should be evaluated in the context of induction and post-
remission chemotherapy before one draws a negative con-
clusion about the value of transplantation.
RITSURO SUZUKI, M.D.
MASAO SETO, M.D.
YASUO MORISHIMA, M.D.
Aichi Cancer Center
Nagoya 464-8681, Japan
1. Zittoun RA, Mandelli F, Willemze R, et al. Autologous or allogeneic
bone marrow transplantation compared with intensive chemotherapy in
acute myelogenous leukemia. N Engl J Med 1995;332:217-23.
2. Burnett AK, Goldstone AH, Stevens RM, et al. Randomised compari-
son of addition of autologous bone-marrow transplantation to intensive
chemotherapy for acute myeloid leukaemia in first remission: results of
MRC AML 10 trial. Lancet 1998;351:700-8.
To the Editor: The study by Cassileth et al. has two major
flaws. The first concerns compliance with the study, its ef-
fect on the intention-to-treat analysis, and the use of the
O’Brien–Fleming boundary. Although 116 patients were
assigned to undergo autologous bone marrow transplanta-
tion, only 63 (54 percent) actually did so. We believe that
for an intention-to-treat analysis, it is unacceptable to retain
in the analysis the 21 patients (18 percent of the group)
who, after randomization, withdrew from the study, when
only 5 patients in the cytarabine group and 3 in the alloge-
neic-transplantation group withdrew. This raises the ques-
tion of a bias of the participating physicians against autolo-
gous bone marrow transplantation. The decision to close
the study when there was no difference between groups in
the rate of leukemia-free survival, the primary end point, is
not allowed when the O’Brien–Fleming boundary is used.
The second flaw concerns cytogenetic features, the most
important independent prognostic factors in patients with
acute myeloid leukemia. Cassileth et al. report only the
distribution of karyotypes to show that there was no dif-
The New England Journal of Medicine
Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
ference between the treatment groups. A Medical Re-
search Council study of acute myeloid leukemia (MRC
AML 10)1 concluded that the value of each therapeutic
strategy depended on cytogenetic features, with worse re-
sults of allogeneic bone marrow transplantation only in
patients with a good prognosis.
We believe that Cassileth et al. convey the misleading
message that any type of bone marrow transplantation is
not advisable in patients with acute myeloid leukemia in
first remission.
NORBERT-CLAUDE GORIN, M.D.
Hôpital Saint-Antoine
75012 Paris, France
MYRIAM LABOPIN, M.D.
European Cooperative Group
for Blood and Marrow Transplantation
75006 Paris, France
1. Burnett AK, Goldstone AH, Stevens RM, et al. Randomised compari-
son of addition of autologous bone-marrow transplantation to intensive
chemotherapy for acute myeloid leukaemia in first remission: results of
MRC AML 10 trial. Lancet 1998;351:700-8.
To the Editor: In his editorial on the role of bone mar-
row transplantation in patients with acute myeloid leuke-
mia in first remission (Dec. 3 issue),1 Dr. Burnett states,
“It already appears justifiable to reserve transplantation for
second-line treatment in patients with a favorable karyo-
type and for children.” We dispute this statement with re-
spect to children. Every large comparative study reported
to date has demonstrated superior survival among children
assigned to undergo allogeneic bone marrow transplanta-
tion.2-5 For example, the Children’s Cancer Group showed
that children and adolescents with HLA-compatible family
donors had significantly longer disease-free survival and
overall survival after a remission had been achieved than
patients with no HLA-compatible donors who were ran-
domly assigned to undergo autologous bone marrow trans-
plantation or chemotherapy.2,3 The compliance rate in this
large trial, in which 656 patients in remission were eligible
for randomization, was 75 percent.
These results confirm those of two previous Children’s
Cancer Group trials involving 800 patients in remission,6,7
as well as the results of the St. Jude,8 Italian Association of
Pediatric Hematology and Oncology (AIEOP), Pediatric
Oncology Group, and Medical Research Council (MRC)
trials.1,4,5 Although we agree that the role of autologous
bone marrow transplantation in first remission remains
questionable, children with HLA-compatible family do-
nors clearly benefit from undergoing allogeneic bone mar-
row transplantation in first remission.
WILLIAM G. WOODS, M.D.
South Carolina Cancer Center
Columbia, SC 29203
JEAN E. SANDERS, M.D.
Fred Hutchinson Cancer Research Center
Seattle, WA 98109-1024
STEVEN NEUDORF, M.D.
Children’s Hospital
Pittsburgh, PA 15213
Vol ume 340 Numb e r 18 · 1437
 The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
1. Burnett AK. Transplantation in first remission of acute myeloid leuke-
mia. N Engl J Med 1998;339:1698-700.
2. Woods WG, Kobrinsky N, Buckley JD, et al. Timed-sequential induc-
tion therapy improves postremission outcome in acute myeloid leukemia:
a report from the Children’s Cancer Group. Blood 1996;87:4979-89.
3. Woods WG, Neudorf S, Gold S, et al. Aggressive post-remission che-
motherapy is better than autologous bone marrow transplantation and al-
logeneic BMT is superior to both in children with acute myeloid leukemia.
Prog Proc Am Soc Clin Oncol 1996;15:368. abstract.
4. Ravindranath Y, Yeager AM, Chang MN, et al. Autologous bone mar-
row transplantation versus intensive consolidation chemotherapy for acute
myeloid leukemia in childhood. N Engl J Med 1996;334:1428-34.
5. Stevens RF, Hann IM, Wheatley K, Gray RG. Marked improvements in
outcome with chemotherapy alone in paediatric acute myeloid leukaemia:
results of the United Kingdom Medical Research Council’s 10th AML tri-
al. Br J Haematol 1998;101:130-40.
6. Nesbit ME, Buckley JD, Feig SA, et al. Chemotherapy for induction of
remission of childhood acute myeloid leukemia followed by marrow trans-
plantation or multiagent chemotherapy: a report from the Childrens Can-
cer Group. J Clin Oncol 1994;12:127-35.
7. Wells RJ, Woods WG, Buckley JD, et al. Treatment of newly diagnosed
children and adolescents with acute myeloid leukemia: a Childrens Cancer
Group Study. J Clin Oncol 1994;12:2367-77.
8. Dahl GV, Kalwinsky DK, Mirro J Jr, et al. Allogeneic bone marrow
transplantation in a program of intensive sequential chemotherapy for chil-
dren and young adults with acute nonlymphocytic leukemia in first remis-
sion. J Clin Oncol 1990;8:295-303.
The authors reply:
To the Editor: We agree with Kanda et al. that therapy
should be guided by prognostic factors, but we believe
that cytogenetic features are more precise prognostic indi-
cators than the FAB classification. Drs. Gorin and Labopin
note that we failed to correlate cytogenetic features with
therapeutic outcomes. The initial version of our manu-
script included such an analysis. It was deleted because the
Journal ’s reviewers believed that the differences observed
among the cytogenetic subgroups1 were of limited reliabil-
ity because of the small number of patients in each sub-
group. We strongly agree that decisions about therapy in
individual patients should be based on an evaluation of the
patient’s prognosis rather than on the results of our study
for the population of patients as a whole.
In our article, we acknowledge the potential flaw cited by
Drs. Gorin and Labopin — namely, that only a fraction of
patients in our study, as in other studies, actually underwent
the assigned autologous bone marrow transplantation. In
our article, we explain the rationale for our intention-to-
treat analysis and describe the biases that influence analyses
that include only patients who actually undergo the proce-
dure. We are aware that the use of the O’Brien–Fleming
boundary involves rules for stopping the study early when
the data support the conclusion that there are significant
treatment differences. As we stated, however, the data mon-
itoring committee of the Eastern Cooperative Oncology
Group directed the study team to unblind the results when
it became clear that the chance of observing a significant
difference in disease-free survival was exceedingly small.
Perez-Calvo and Brugarolas suggest that the results of this
study, in which chemotherapy-purged bone marrow was
transplanted, should not be compared with the results of
studies in which unpurged bone marrow was used.2-4 Only
two of the studies they cite2,4 reported a benefit in terms of
disease-free survival after autologous bone marrow trans-
plantation as compared with chemotherapy; overall survival
was not improved in two of the studies2,3 and was only mar-
ginally improved in the other study.4 The lack of substantive
improvement in survival reported in these studies is consis-
1438 · May 6 , 19 9 9
The New England Journal of Medicine
Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
tent with our results. Perez-Calvo and Brugarolas also be-
lieve that the mortality rate of 14 percent in our group of
patients who underwent autologous transplantation with
purged bone marrow adversely affected the disease-free sur-
vival and overall survival of this group as a whole. This may
be true and, as we noted, may suggest a way to improve the
results of autologous bone marrow transplantation, with the
use, for example, of peripheral-blood stem cells.
The caveat of Suzuki et al. is well taken. Our results can
be applied only to patients who receive the induction and
consolidation therapy we used. Other approaches to pre-
remission and postremission therapy, preparative regimens
for transplantation, and stem-cell transplantation may have
different results. New therapeutic strategies need to be
compared with conventional therapy in randomized studies
that analyze the data according to prognostic subgroups.
PETER A. CASSILETH, M.D.
University of Miami Sylvester Comprehensive Cancer Center
Miami, FL 33136
FREDERICK R. APPELBAUM, M.D.
Fred Hutchinson Cancer Research Center
Seattle, WA 98109-1024
PETER H. WIERNIK, M.D.
Our Lady of Mercy Comprehensive Cancer Center
Bronx, NY 10466
1. Slovak ML, Kopecky KJ, Cassileth PA, et al. Karyotypic analysis predicts
outcome of pre- and post-remission therapy in adult acute myeloid leuke-
mia (AML): a SWOG/ECOG intergroup study. Blood 1998;92:Suppl 1:
678a. abstract.
2. Zittoun RA, Mandelli F, Willemze R, et al. Autologous or allogeneic
bone marrow transplantation compared with intensive chemotherapy in
acute myelogenous leukemia. N Engl J Med 1995;332:217-23.
3. Harousseau JL, Cahn JY, Pignon B, et al. Comparison of autologous
bone marrow transplantation and intensive chemotherapy as postremission
therapy in adult acute myeloid leukemia. Blood 1997;90:2978-86.
4. Burnett AK, Goldstone AH, Stevens RM, et al. Randomised compari-
son of addition of autologous bone-marrow transplantation to intensive
chemotherapy for acute myeloid leukaemia in first remission: results of
MRC AML 10 trial. Lancet 1998;351:700-8.
To the Editor: Any benefit of bone marrow transplanta-
tion for acute myeloid leukemia depends on how effective
the comparative chemotherapy is. In the AIEOP trial,1 the
chemotherapy group had a poorer outcome (rate of dis-
ease-free survival, 27 percent) than that expected with
contemporary treatment in children. Improvements in
chemotherapy continue to improve the results, as report-
ed by Dr. Woods and colleagues on behalf of the Chil-
dren’s Cancer Group2 and by the United Kingdom MRC
Paediatric Leukaemia Working Party, which recently re-
ported a 56 percent rate of overall survival at seven years
in recipients of chemotherapy alone; in fact, this study
showed no superiority of transplantation.3 When the data
in the MRC study were “time-adjusted” so that survival
was calculated from day 155 of remission, which was the
median elapsed time between the induction of clinical re-
mission and transplantation, 66 percent of the patients
who received chemotherapy alone (range, 58 to 74 per-
cent) were alive at seven years, as compared with 73 per-
cent of those who underwent transplantation (range, 62
to 84 percent). The overall rate of survival at seven years
did not differ significantly between the patients with do-
 C ORR ES POND ENCE
nors and those without donors (70 percent and 60 per-
cent, respectively; P=0.1).
Even when children with favorable karyotypes were ex-
cluded from the MRC analysis, there was no advantage of
bone marrow transplantation. In the MRC study of acute
myeloid leukemia (MRC AML 10), from which this experi-
ence was derived, the rate of second remission among chil-
dren who had relapses after chemotherapy was 90 percent,
and the survival rate among those with relapses was 34 per-
cent. Given that this is the only time-adjusted comparison of
children with donors and those without donors that has been
reported so far and that it showed no significant advantage
of transplantation with respect to survival, it is important
to take into account the associated morbidity and possible
late adverse effects, as well as the fact that these children have
an excellent chance of a second remission. Under these cir-
cumstances, it does seem justifiable to delay transplantation.
This approach has been adopted by the MRC Paediatric
Leukaemia Working Party in its current trial, which is show-
ing a further improvement in survival. The results of the
Children’s Cancer Group study 2 and the Pediatric Oncolo-
gy Group study 4 may indeed conflict with the MRC data
when a time-adjusted comparison of children with donors
and those without donors is performed, in which case the
value of allogeneic transplantation is uncertain and should
continue to be examined in the context of a prospective trial.
ALAN K. BURNETT, M.D.
University of Wales College of Medicine
Cardiff CF4 4XN, United Kingdom
1. Amadori S, Testi AM, Arico M, et al. Prospective comparative study of
bone marrow transplantation and postremission chemotherapy for child-
hood acute myelogenous leukemia. J Clin Oncol 1993;11:1046-54.
2. Woods WG, Neudorf S, Gold S, et al. Aggressive post-remission che-
motherapy is better than autologous bone marrow transplantation and
allogeneic BMT is superior to both in children with acute myeloid leuke-
mia. Prog Proc Am Soc Clin Oncol 1996;15:368. abstract.
3. Stevens RF, Hann IM, Wheatley K, Gray RG. Marked improvements in
outcome with chemotherapy alone in paediatric acute myeloid leukemia:
results of the United Kingdom Medical Research Council’s 10th AML
trial. Br J Haematol 1998;101:130-40.
4. Ravindranath Y, Yeager AM, Chang MN, et al. Autologous bone mar-
row transplantation versus intensive consolidation chemotherapy for acute
myeloid leukemia in childhood. N Engl J Med 1996;334:1428-34.
Long-Acting Gonadotropin-Releasing Hormone
Implant to Maintain Medical Castration for Two
Years in Men with Prostate Cancer
To the Editor: Medical castration with long-acting go-
nadotropin-releasing hormone agonists is an effective ther-
apy for men with metastatic prostate cancer.1,2 Initially,
these agonists had to be given daily, but depot prepara-
tions that can be administered at intervals of one to three
months are now available.3,4 We describe here the response
to a longer-acting implant releasing the gonadotropin-
releasing hormone agonist histrelin (Hydron, Hydro Med
Sciences) in men with prostate cancer.
We treated 13 men 65 to 88 years of age who had pros-
tate cancer that extended beyond the prostate capsule with
50 mg of the gonadotropin-releasing hormone agonist his-
trelin, administered with a cylindrical reservoir drug-release
device 30 mm long and 3.5 mm in diameter, inserted sub-
cutaneously in the upper arm with local anesthesia. The in
The New England Journal of Medicine
Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
vitro rate of histrelin release from this device is approxi-
mately 60 µg daily for at least one year. Five men received
one implant, and the remaining eight received two implants.
To prevent the transient increase in the secretion of lu-
teinizing hormone and testosterone when gonadotropin-
releasing hormone agonist therapy is begun1,2 — a response
known as the flare phenomenon — the men received an an-
tiandrogen starting two weeks before insertion of the im-
plant and continuing for three months thereafter. Nine men
received flutamide, and four cyproterone acetate. During
therapy, blood samples were collected for measurements of
serum luteinizing hormone, testosterone, and prostate-spe-
cific antigen and radionuclide bone scanning was repeated
every six months. The study was approved by the Human
Investigation Committee of Shaare Zedek Medical Center,
and all the men gave written informed consent.
Serum luteinizing hormone was measured by enzyme
immunoassay (Abbott Laboratories), testosterone was meas-
ured by radioimmunoassay (Diagnostic Products), and
prostate-specific antigen was measured by an immunoradi-
ometric assay (BYK, Sangtec Diagnostica). The lower lim-
its of sensitivity for the luteinizing hormone and testoster-
one assays were 0.1 IU per liter and 10 ng per deciliter
(0.35 nmol per liter), respectively.
Serum prostate-specific antigen concentrations in these
men before treatment ranged from 29 to 931 ng per mil-
liliter (normal value, «4). The values began to decrease
during antiandrogen therapy and continued to fall there-
after, reaching values of <4 ng per milliliter in all the men
after three months. In all the men, serum luteinizing hor-
mone and testosterone concentrations were undetectable
after 1 month of treatment and remained so after 17 to
24 months.
Three men had hot flashes, but there were no other ad-
verse effects. Eleven of the 13 men remain in remission,
with serum prostate-specific antigen concentrations in the
normal range, but the concentrations of serum prostate-
specific antigen began to rise after 11 months of treatment
in 2 men. This histrelin implant is effective in maintaining
the suppression of luteinizing hormone and testosterone
secretion for at least two years in men with prostate cancer.
IRVING M. SPITZ, M.D.
BORIS CHERTIN, M.D.
TZINA LINDENBERG, M.A.
AMICUR FARKAS, M.D.
Shaare Zedek Medical Center
Jerusalem 91031, Israel
1. Walsh PC. Physiologic basis for hormonal therapy in carcinoma of the
prostate. Urol Clin North Am 1975;2:125-40.
2. Labrie F. Endocrine therapy of prostate cancer: optimal form and tim-
ing. J Clin Endocrinol Metab 1995;80:1066-71.
3. Crawford ED, Eisenberger MA, McLeod DG, et al. A controlled trial
of leuprolide with and without flutamide in prostatic carcinoma. N Engl J
Med 1989;321:419-24. [Erratum, N Engl J Med 1989;321:1420.]
4. Dijkman GA, del-Moral PF, Plasman JW, et al. A new extra long acting
depot preparation of the LHRH analogue Zoladex: first endocrinological
and pharmacokinetic data in patients with advanced prostate cancer. J Ster-
oid Biochem Mol Biol 1990;37:933-6.
Unruptured Intracranial Aneurysms
To the Editor: The conclusions reached by Wiebers et al.
(Dec. 10 issue)1 about the natural history of aneurysmal
Vol ume 340 Numb e r 18 · 1439
 The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
TABLE 1. DISTRIBUTION OF ANEURYSMS.
TYPE LOCATION* GROUP 1 GROUP 2
no. of aneurysms (%)
Unprotected in sub- ACA, MCA, 570 (58) 698 (73)
arachnoid space PCA, VB
Protected from sub- CCA, ICA 407 (42) 262 (27)
arachnoid space
Total no. of aneurysms 977 960
*ACA denotes anterior communicating artery, MCA mid-
dle cerebral artery, PCA posterior communicating artery, VB
vertebrobasilar arteries, CCA cavernous carotid artery, and
ICA internal carotid artery.
subarachnoid hemorrhage arouse concern because of the
inhomogeneous grouping of patients and the differences
in the rates of subarachnoid hemorrhage at different sites.
Since they included intracavernous aneurysms in their de-
termination of rupture rates, the rates for aneurysms truly
within the subarachnoid space may be underestimated,
since intracavernous aneurysms rarely cause subarachnoid
hemorrhage.2 To a lesser extent, the same caveat may apply
to proximal infraclinoid ophthalmic aneurysms and those
arising from the clinoidal segment of the carotid artery,
since they are often protected by dura and bone. The
grouping of aneurysms in these locations with aneurysms
that are free within the subarachnoid space combines cat-
egories of aneurysms that entail very different risks of
hemorrhage. Of the 1937 aneurysms, 669 (34.5 percent)
were in these proximal locations, a point that may have im-
portant implications for the overall findings (Table 1).
From the data provided, we cannot determine the effect
of this distribution of aneurysms on the much higher (by
a factor of 11) risk of subarachnoid hemorrhage among the
patients in group 2. Certainly, the difference of 15 percent-
age points between the two groups in the proportion of
aneurysms free in the subarachnoid space may account for
some of the results. The data shown in Table 1 of the arti-
cle by Wiebers et al. suggest that the incidence of single an-
eurysms was the same in the two groups. Since patients in
group 2 had had a previous subarachnoid hemorrhage that
had been treated, they represent a population with a higher
rate of multiple aneurysms and are not comparable to the
patients in group 1. This factor may also have contributed
to the difference in the natural history of unruptured intra-
cranial aneurysms between the groups and must be ac-
knowledged in any discussion of the natural history.
Another concern of ours is the relation between sub-
arachnoid hemorrhage and the size of the aneurysm. Al-
though Wiebers et al.3 have written about the absence of
subarachnoid hemorrhage in patients with aneurysms that
are less than 10 mm in diameter, many large series have
shown that the mean diameter of aneurysms in patients
who present with subarachnoid hemorrhage is less than
10 mm, as noted by Dr. Caplan in his editorial.4 There-
fore, we conclude that there is little assurance that an an-
eurysm that is less than 10 mm will not bleed.
We believe that the conclusions of this article should be
modified to reflect a higher risk of subarachnoid hemor-
1440 · May 6 , 19 9 9
The New England Journal of Medicine
Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
rhage at certain sites, since this will affect the manner in
which treatment options are presented to patients.
ALEJANDRO BERENSTEIN, M.D.
EUGENE S. FLAMM, M.D.
MARK J. KUPERSMITH, M.D.
Beth Israel Medical Center
New York, NY 10128
1. The International Study of Unruptured Intracranial Aneurysms Investi-
gators. Unruptured intracranial aneurysms — risk of rupture and risks of
surgical intervention. N Engl J Med 1998;339:1725-33.
2. Kupersmith MJ, Hurst R, Berenstein A, Choi IS, Jafar J, Ransohoff J.
The benign course of cavernous carotid artery aneurysms. J Neurosurg
1992;77:690-3.
3. Wiebers DO, Whisnant JP, O’Fallon WM. The natural history of unrup-
tured intracranial aneurysms. N Engl J Med 1981;304:696-8.
4. Caplan LR. Should intracranial aneurysms be treated before they rup-
ture? N Engl J Med 1998;339:1774-5.
To the Editor: Although we were encouraged to see a re-
port on the important problem of unruptured intracranial
aneurysms, we are concerned that selection bias, especially
in the retrospective cohort, may have undermined the
study’s findings and recommendations.
We were approached to participate in this study and in-
clude any patients with unruptured intracranial aneurysms
who were being followed. Although we had several such pa-
tients, we declined the invitation because of our concern
that the cohort would not be representative of the overall
population of patients with unruptured aneurysms. For in-
stance, the vast majority of patients with unruptured aneu-
rysms who are referred to our institution have already been
surgically treated, leaving only a small minority with aneu-
rysms with a low risk on the basis of the natural history, and
these aneurysms were, for example, heavily calcified, partly
intracavernous, tiny and laterally located, or in elderly pa-
tients with other medical problems. Unfortunately, aneu-
rysms such as these also often pose a greater surgical risk,
further complicating the issue of their inclusion in a study.
We think that the data reported confirm our concern
about selection bias. The cohort consisted of 1449 pa-
tients at 53 centers who were given a diagnosis over a pe-
riod of 21 years (1970 to 1991). All the centers are re-
gional referral institutions that would be expected, on the
basis of the volume at our institution, to see 60 to 80 pa-
tients with newly diagnosed unruptured aneurysms annu-
ally. Assuming this to be the case, outcome data would
have been reported for only 1.3 patients per year (2 per-
cent of all patients seen). Even if one assumes the volume
in the referral centers to be only 10 percent of this value,
the cohort would represent only 20 percent of the patients
in these centers, which is still too small a percentage to be
representative of the population at large.
Thus, although we applaud any plans for a true popula-
tion-based assessment and some type of randomized trial
addressing this problem and hope that this article will trigger
enough controversy to see that goal accomplished, we recall
the initial impressions that carotid artery disease was benign
and not in need of treatment, which were reversed by defin-
itive studies showing that intervention was appropriate for
patients who were properly identified. The same could apply
to unruptured aneurysms, which the study shows still have
a case fatality rate of 66 percent when they are left to bleed.
 C ORR ES POND ENCE
For those of us treating this disease, the identification
of appropriate surgical candidates, combined with more
cost-effective screening, probably offers the greatest hope,
since ongoing efforts to prevent early rebleeding, cure vas-
ospasm, reverse severe brain damage, and refine microsur-
gical and endovascular techniques are having little effect
on overall morbidity due to this disease.
E. SANDER CONNOLLY, JR., M.D.
J.P. MOHR, M.D.
ROBERT A. SOLOMON, M.D.
Columbia University
New York, NY 10032
To the Editor: The annual rupture rate reported by Wie-
bers et al. is considerably lower than that reported in other
respected studies.1,2 One study found that unruptured an-
eurysms smaller than 5 mm in diameter that subsequently
ruptured were larger after rupture.3 The Aneurysmal Sub-
arachnoid Hemorrhage Cooperative Study2 demonstrated
that 24 percent of ruptured aneurysms are 5 mm or less.
Rupture of aneurysms smaller than 10 mm is not uncom-
mon in our practice. The low rate of rupture of small an-
eurysms reported by Wiebers et al. is most likely explained
by the fact that growth and rupture are time-dependent.
Follow-up was not long enough to allow the expansion
and rupture of aneurysms.
The rates of surgical morbidity and mortality were high-
er in this study than in other series.4 There have been mi-
croneurosurgical advances since 1970, when the retrospec-
tive component began. It would not be accurate to use
high complication rates when one is analyzing the risks
and benefits of the obliteration of aneurysms. We are sure
that the authors accept the seriousness of these lesions and
the potential for poor outcomes (a mortality rate of up to
50 percent after rupture). In addition, in 42 of the 205
patients who died during the 7.5 years of follow-up, death
was caused by intracranial hemorrhage. Was this due to
the aneurysm? Were autopsy data available?
Patients in whom the aneurysm was manipulated within
30 days after diagnosis were excluded. This cohort most
likely represents younger patients who have aneurysms
with worrisome features that make the decision to operate
straightforward. If these assumptions are true, the rate of
surgical complications would be lower in this group. The
number of such patients and the reasons for treatment are
not reported.
The locations of the aneurysms in this study differ from
those in previous studies and suggest that there may have
been a location-specific bias.5 There were 256 cavernous
aneurysms. The likelihood of rupture is low if the lesion is
completely intracavernous. These lesions account for 18
percent of aneurysms in the retrospective study, but for
only 6 percent of cases of subarachnoid hemorrhage.
To advise patients with unruptured aneurysms properly,
three factors require consideration: the size and location
of the aneurysm and the patient’s age. The study does not
provide information regarding the rates of surgical compli-
cations as a function of the location or the size of the an-
eurysm or the patient’s age. Unfortunately, all patients are
combined into one group that reflects various micro-
neurosurgical techniques. This approach is very mislead-
The New England Journal of Medicine
Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
ing. We hope that this report will prompt further investi-
gation into this important clinical issue.
PHILIP E. STIEG, PH.D., M.D.
ROBERT FRIEDLANDER, M.D.
Brigham and Women’s Hospital
Boston, MA 02115
1. Juvela S, Porras M, Heiskanen O. Natural history of unruptured intra-
cranial aneurysms: a long-term follow-up study. J Neurosurg 1993;79:174-
82.
2. Sahs AL, Nibbelink DW, Torner JC. Aneurysmal subarachnoid hemor-
rhage: report of the Cooperative Study. Baltimore: Urban & Schwarzen-
berg, 1981.
3. Yasui N, Magarisawa S, Suzuki A, Nishimura H, Okudera T, Abe T.
Subarachnoid hemorrhage caused by previously diagnosed, previously un-
ruptured intracranial aneurysms: a retrospective analysis of 25 cases. Neu-
rosurgery 1996;39:1096-100.
4. King JT Jr, Berlin JA, Flamm ES. Morbidity and mortality from elective
surgery for asymptomatic, unruptured, intracranial aneurysms: a meta-
analysis. J Neurosurg 1994;81:837-42.
5. McCormick WF. Vascular disorders of nervous tissue: anomalies, mal-
formations, and aneurysms. In: Bourne GH, ed. The structure and func-
tion of nervous tissue. Vol. 3. Biochemistry and disease. New York: Aca-
demic Press, 1969:537-95.
To the Editor: In his editorial, Dr. Caplan states that “pa-
tients with previously ruptured aneurysms had 11 times
the rate of rupture of patients without prior hemorrhage.”
As I interpret the data of Wiebers et al., this excess is ap-
plicable only to the subgroup of patients with aneurysms
that were less than 10 mm in diameter. Among patients
with aneurysms that were 10 mm or more, the rates of
rupture are roughly similar in the group with prior hem-
orrhage and the group without it.
ALLAN BRETT, M.D.
University of South Carolina School of Medicine
Columbia, SC 29203
The authors reply:
To the Editor: Approximately half the patients with cav-
ernous aneurysms also had noncavernous unruptured an-
eurysms. Although intracavernous aneurysms can cause
subarachnoid hemorrhage, most are protected from the
subarachnoid space. Other internal carotid aneurysms are
not protected. The annual rates of rupture are only slightly
increased by the exclusion of patients with cavernous an-
eurysms (Table 1); patients in group 2 who had small (<10
mm) unruptured aneurysms remain approximately 10
times as likely to have a subsequent rupture as patients
with small aneurysms in group 1.
Currently, the numbers of patients are too small (only
a total of 32 aneurysmal ruptures) to determine meaning-
ful rupture rates in group 1 and group 2 according to the
six different locations and two size categories (24 sub-
groups). The presence of multiple unruptured intracranial
aneurysms was not a predictor of future rupture independ-
ent of the size and location of the largest unruptured an-
eurysm. Unruptured aneurysms were more likely to rup-
ture in patients in group 2, but this risk does not appear
to be related to the presence of multiple unruptured intra-
cranial aneurysms.
The concern about microneurosurgical advances that have
occurred since 1970 and the exclusion of patients who un-
Vol ume 340 Numb e r 18 · 1441
 The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne

1. Menghini VV, Brown RD Jr, Sicks JD, O’Fallon WM, Wiebers DO. The
TABLE 1. MEAN ANNUAL RATES OF incidence and prevalence of intracranial saccular aneurysms and hemorrhage
in Olmsted County, Minnesota, 1965 to 1995. Neurology 1998;51:405-11.
CONFIRMED SUBARACHNOID HEMORRHAGE.

ALL PATIENTS EXCEPT To the Editor: Dr. Brett is correct. Aneurysms that were
DIAMETER OF ALL THOSE WITH CAVERNOUS less than 10 mm in diameter were 11 times as likely to rup-
UNRUPTURED ANEURYSM PATIENTS ANEURYSMS ture in patients who had prior bleeding from other aneu-
percent per year rysms as in patients without prior subarachnoid hemor-
rhage. Aneurysms that were at least 10 mm had a similar
Group 1 (no subarachnoid rate of rupture whether or not there had been prior sub-
hemorrhage)
<10 mm 0.05 0.07 arachnoid hemorrhage.
»10 mm 0.95 1.38 LOUIS R. CAPLAN, M.D.
Group 2 (subarachnoid
hemorrhage) Beth Israel Deaconess Medical Center
<10 mm 0.56 0.60 Boston, MA 02215
»10 mm 0.64 0.72

Rupture of Cerebral Aneurysm

derwent surgery in the first 30 days after diagnosis is not rel-
evant, since operative morbidity and mortality were assessed
only in the patients in the prospective group, beginning in
late 1991. The base-line characteristics of the surgically treat-
ed patients and those who were not so treated were virtually
identical, with minor differences in the mean age and aneu-
rysmal size only among patients in group 1, underscoring
the lack of consensus about the selection of patients with
unruptured intracranial aneurysms as candidates for surgery.
Comparison of our retrospective results with those of a
30-year study (1965 to 1995) in Rochester, Minnesota, of
a population-based sample of patients with intracranial
aneurysms1 (and unpublished data) yielded strikingly sim-
ilar demographic characteristics, aneurysmal characteris-
tics, and associated medical conditions, suggesting that
our group (representing approximately 40 percent of pa-
tients with unruptured aneurysms who were seen at par-
ticipating centers) may be representative of the general
population. A randomized trial would involve a much
greater prospective selection bias than our study because
many patients would refuse to participate.
Data on the natural history of unruptured intracranial
aneurysms confirm the conclusion that judgments about
the probability of the rupture of such aneurysms cannot
be extrapolated from data on patients with ruptured aneu-
rysms. It appears that most aneurysms that are going to
rupture do so when they form or soon afterward and that
the critical size in terms of rupture is smaller for aneu-
rysms that rupture early.
In our cohort, the patient’s age significantly predicted
operative morbidity and mortality. The influences of the
size and location of aneurysms in this study are not yet
clear because of insufficient numbers of patients; this is
one of the central reasons that the study has been extend-
ed to involve a total of 5500 patients.
DAVID O. WIEBERS, M.D.
DAVID G. PIEPGRAS, M.D.
JOHN HUSTON III, M.D.
Mayo Clinic
Rochester, MN 55905
FOR THE INTERNATIONAL STUDY OF UNRUPTURED
INTRACRANIAL ANEURYSMS INVESTIGATORS
during Angiography
To the Editor: Dr. Stockinger’s contribution to Images
in Clinical Medicine (Dec. 10 issue)1 shows the fatal peri-
angiographic rupture of an aneurysm of the posterior in-
ferior cerebellar artery in a 35-year-old woman. Forceful
injection of contrast medium into a vertebral artery to vis-
ualize the contralateral posterior inferior cerebellar artery
through retrograde opacification of the contralateral ver-
tebral artery is a procedure that has long been used in or-
der to avoid the need for selective catheterization of both
vertebral arteries during a workup for subarachnoid hem-
orrhage. It must be remembered, however, that the intra-
aneurysmal pressure is increased by the injection of con-
trast medium,2 and that the rise in pressure is correlated
with the rate and volume of the injection.3
The maneuver thus appears to be inappropriate. It
should be reserved for the small fraction of generally older
patients with atheromatous vessels that preclude the safe,
selective study of one vertebral artery, and then only after
an ipsilateral aneurysm has been reasonably ruled out by a
gentle injection of contrast medium. The injection of 6 ml
of nonionic contrast material at a rate of 3 ml per second
is generally sufficient to obtain an image of excellent quality
with modern digital-subtraction angiographic equipment.
PHILIPPE GAILLOUD, M.D.
KIERAN J. MURPHY, M.D.
Johns Hopkins Hospital
Baltimore, MD 21287
1. Stockinger Z. Rupturing aneurysm of the posterior inferior cerebellar
artery. N Engl J Med 1998;339:1758.
2. Sorimachi T, Takeuchi S, Koike T, Minakawa T, Tanaka R. Intra-aneu-
rysmal pressure changes during angiography in coil embolization. Surg
Neurol 1997;48:451-7.
3. Saitoh H, Hayakawa K, Nichimura K, et al. Intracarotid blood pressure
changes during contrast medium injection. AJNR Am J Neuroradiol 1996;
17:51-4.
The editor replies:
The points made by Drs. Gailloud and Murphy are well
taken. Unfortunately, Dr. Stockinger is unable to supply
any additional information about the procedure.
JEROME P. KASSIRER, M.D.

1442 · May 6 , 19 9 9

The New England Journal of Medicine

Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
 C ORR ES POND ENCE
Parkinsonism after Taking Ecstasy
To the Editor: Recreational use of 3,4-methylenedioxy-
methamphetamine (MDMA, or “ecstasy”), a hallucinogen,
has increased both in Europe and the United States.1 This
substance is manufactured in illicit laboratories from a vari-
ety of organic ketone precursors. MDMA, which is struc-
turally related both to the stimulant amphetamine and to
the hallucinogen mescaline, promotes the release of both
serotonin and dopamine from synaptic terminals.2 We re-
port a case of parkinsonism after repeated use of this drug.
A 29-year-old man had slight clumsiness of his upper and
lower extremities in August 1998. During the following
four weeks he began to have difficulty walking and lost the
ability to write and to drive. He was unable to continue his
work in retail merchandising and then could not live inde-
pendently. The results of magnetic resonance imaging, elec-
troencephalography, lumbar puncture, and positron-emis-
sion tomography with [18F]fluorodeoxyglucose were normal.
Tests for antibodies to the human immunodeficiency virus
were negative on two occasions. Eleven weeks after the on-
set of symptoms, examination revealed a disturbance in gait
and impairment of fine coordination. The patient’s condi-
tion continued to worsen, and eight weeks later examina-
tion revealed bradykinesia of the face and limbs, absence of
blinking, hypokinesia in relation to speech, postural insta-
bility, and a markedly parkinsonian gait. Cognitive function
was normal, and there was no tremor.
The patient had ingested MDMA nine times in 1997
and once more in May 1998. He denied having used any
other illicit substances except cannabis. Treatment with
the maximal tolerated doses of levodopa and pramipexole
did not improve his symptoms.
Parkinsonism has not previously been associated with the
use of MDMA. Although we have no firm evidence of a
causal relation between this patient’s drug use and his par-
kinsonism, there are no other tenable explanations. Idio-
pathic Parkinson’s disease rarely develops in this age group
and is responsive to medication, and neither it nor parkin-
sonism of any other neurodegenerative cause progresses this
rapidly. Our patient’s condition most closely resembled the
parkinsonism produced by 1-methyl-4-phenyl-1,2,3,6-tet-
rahydropyridine, a byproduct of the illicit manufacture of
meperidine, which is highly toxic to neurons of the substan-
tia nigra. Positron-emission tomographic studies have shown
that even subclinical exposure to this substance produces
progressive nigrostriatal damage, making the subject vulner-
able to subsequent parkinsonism.3
We hypothesize that our patient may have had parkin-
sonism as a result of a delayed neurotoxic effect of MDMA
on the substantia nigra and striatum. There is evidence for
such dopaminergic neurotoxicity in animals,4,5 and delayed
neurotoxicity in humans could occur as a result of neuro-
nal damage induced by free radicals.5 Recent research re-
garding the neurotoxic effects of this substance has fo-
cused on serotonergic neurotoxicity and impairment of
memory.6 This patient’s case suggests that the effects of
MDMA on dopaminergic systems should be studied fur-
ther to characterize more fully the dangers of its use.
SCOTT MINTZER, M.D.
SUSAN HICKENBOTTOM, M.D.
SID GILMAN, M.D.
University of Michigan
Ann Arbor, MI 48109
The New England Journal of Medicine
Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
1. Johnston LD, O’Malley PM, Bachman JG. National survey results on
drug use from the Monitoring the Future Study, 1975–1995. Vol. 2. Col-
lege students and young adults. Rockville, Md.: National Institutes on
Drug Abuse, 1997.
2. Steele TD, McCann UD, Ricaurte GA. 3,4-Methylenedioxymetham-
phetamine (MDMA, “Ecstasy”): pharmacology and toxicology in animals
and humans. Addiction 1994;89:539-51.
3. Vingerhoets FJ, Snow BJ, Tetrud JW, Langston JW, Schulzer M, Calne
DB. Positron emission tomographic evidence for progression of human
MPTP-induced dopaminergic lesions. Ann Neurol 1994;36:765-70.
4. Obradovic T, Imel KM, White SR. Repeated exposure to methylene-
dioxymethamphetamine (MDMA) alters nucleus accumbens neuronal re-
sponses to dopamine and serotonin. Brain Res 1998;785:1-9.
5. Seiden LS, Sabol KE. Methamphetamine and methylenedioxymeth-
amphetamine neurotoxicity: possible mechanisms of cell destruction.
In: Majewska MD, ed. Neurotoxicity and neuropathology associated with
cocaine abuse. NIDA research monograph 163. Washington, D.C.:
Government Printing Office, 1996. (DHHS publication no. (ADM)
96-4019.)
Carvedilol
To the Editor: In his review of carvedilol, Frishman (Dec.
10 issue)1 did not adequately distinguish the effects of the
drug in patients with dilated cardiomyopathy from those in
patients with ischemic cardiomyopathy. Bristow et al.2 have
shown that there are differences between these two disor-
ders in b-adrenergic activation. The Australia–New Zealand
Heart Failure Research Collaborative Group3,4 has assessed
the efficacy of carvedilol in patients with ischemic cardio-
myopathy. The investigators reported an increase in symp-
toms of congestive heart failure with carvedilol as compared
with placebo. Carvedilol did result in an increase in left
ventricular ejection fraction and a decrease in ventricular
size, but both of these effects may have been due to the
lower heart rate and lower blood pressure in these relatively
healthy patients. In addition, neither maximal nor submax-
imal exercise performance changed with use of the drug.
These results argue against the efficacy of carvedilol for re-
lieving symptoms of congestive heart failure in patients
with ischemic cardiomyopathy.
LAWRENCE J. COHN, M.D.
45 E. 72nd St.
New York, NY 10021
1. Frishman WH. Carvedilol. N Engl J Med 1998;339:1759-65.
2. Bristow MR, Anderson FL, Port JD, et al. Differences in beta-adrener-
gic neuroeffector mechanisms in ischemic versus idiopathic dilated cardio-
myopathy. Circulation 1991;84:1024-39.
3. The Australia–New Zealand Heart Failure Research Collaborative
Group. Effects of carvedilol, a vasodilator-beta-blocker, in patients with
congestive heart failure due to ischemic heart disease. Circulation 1995;92:
212-8.
4. Idem. Randomised, placebo-controlled trial of carvedilol in patients
with congestive heart failure due to ischaemic heart disease. Lancet 1997;
349:375-80.
Dr. Frishman replies:
To the Editor: Dr. Cohn suggests that the benefit of
carvedilol may vary according to the cause of heart failure.
In particular, he suggests that patients with ischemic car-
diomyopathy may respond less well than others, citing the
Australia–New Zealand study to support that position.1,2
In the Australia–New Zealand study, symptoms of heart
failure did increase in slightly more patients in the
carvedilol group at 6 months, though not at 12 months.
Vol ume 340 Numb e r 18 · 1443
 The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
The authors questioned whether this represented an effect
of treatment on symptoms of heart failure or simply typi-
cal effects of treatment with a b-adrenergic–antagonist
drug (e.g., fatigue). Regarding the relation between im-
provement in ejection fraction and reductions in heart rate
and blood pressure, the smaller ventricular dimensions
noted suggested that the effect of carvedilol on ejection
fraction was due not only to a decrease in the heart rate
but also to improvement in myocardial contractility. It must
be kept in mind, however, that the benefits of b-adrener-
gic blockade in patients with heart failure are related pri-
marily to a slowing of disease progression and a reduction
in the attendant morbidity and mortality. In this trial,
there was a lower overall risk of serious clinical events re-
sulting in death or hospitalization and thus an estimated
prevention each year of one serious event among every 12
or 13 patients who received carvedilol.
WILLIAM H. FRISHMAN, M.D.
New York Medical College
Valhalla, NY 10595
1. The Australia–New Zealand Heart Failure Research Collaborative
Group. Effects of carvedilol, a vasodilator-beta-blocker, in patients with
congestive heart failure due to ischemic heart disease. Circulation 1995;92:
212-8.
2. Idem. Randomised, placebo-controlled trial of carvedilol in patients
with congestive heart failure due to ischaemic heart disease. Lancet 1997;
349:375-80.
Transient Global Amnesia at High Altitude
To the Editor: Transient focal neurologic conditions at
high altitude without cerebral edema are uncommon but
alarming. We describe two cases of isolated, transient
global amnesia at very high altitude.
The first patient, a healthy 21-year-old man with no his-
tory of migraine, flew to Mexico City (elevation, 2241 m).
He ascended by vehicle to 3950 m over a period of four
days to begin an ascent of Popocatépetl (elevation, 5452 m).
After a day of rest, he climbed to an altitude of 4400 m
without problems. The next morning, the climber appeared
startled and confused. Though alert, he was unable to re-
call the date, the location, or his purpose for being on a
volcano. He was able to state his name, however, and after
prompting, was able to recognize his companions. The re-
sults of an examination were normal, with no ataxia or
dyspnea at rest. A descent of 450 m brought gradual res-
olution of the amnesia. During the next six days, the climb-
er ascended two other volcanoes higher than 5200 m
©1999, Massachusetts Medical Society.
1444 · May 6 , 19 9 9
The New England Journal of Medicine
Downloaded from nejm.org at MCMASTER UNIVERSITY HEALTH SCI LIB on February 23, 2015. For personal use only. No other uses without permission.
Copyright © 1999 Massachusetts Medical Society. All rights reserved.
without problems. At a one-year follow-up examination,
he was in good health.
The second patient, a healthy 60-year-old man with no
history of migraine, dementia, or cerebral vascular disease,
drove to an altitude of 1650 m and hiked to 3000 m to
begin a solo ascent of Mount Rainier (4422 m). At dawn
the next day, he continued climbing to 3760 m, where he
was later found resting. The climber was alert and able to
state his name, but he was unable to recall the date, his
home address, or the current U.S. president. Results of an
examination were normal, with no ataxia or dyspnea at
rest. After a 760-m descent, he became fully oriented, and
the amnesia resolved. At a two-month follow-up inter-
view, he reported good health.
Transient global amnesia is a syndrome of severe, forget-
ful confusion characterized by retrograde memory loss
and total disorientation except for self-identity, followed
by gradual resolution. Commonly, the presumed cause is
transient ischemia.1 The relative hypoxia associated with
exposure to very high altitudes may precipitate transient
global amnesia in some persons. Memory is critically served
by the limbic cortex and involves the temporal hippocam-
pal axis, and even moderate hypoxia due to high altitude
may impair memory.2
Increased intracerebral pressure caused by cerebral ede-
ma at high altitude could have resulted in the amnesia, but
it is unlikely as an isolated finding, and neither of the pa-
tients we describe had evidence of severe altitude illness.
The amnesic episodes developed in association with rapid
ascent to very high altitudes and resolved during descent,
suggesting that exposure to high altitude may have precip-
itated or caused the amnesic events.
Persons with neurologic conditions at high altitudes
benefit from immediate descent, supplemental oxygen, and
possibly carbon-dioxide rebreathing if there are no signs
of increased intracerebral pressure. Patients with persistent
symptoms after descent require prompt evacuation and
thorough evaluation.3
JAMES A. LITCH, M.D., D.T.M.H.
RACHEL A. BISHOP, M.B., CH.B., D.T.M.H.
Kunde Hospital
Kathmandu, Nepal
1. Plum F. Disorders of the cerebral hemispheres and higher brain func-
tions. In: Berkow R, Fletcher AJ, eds. The Merck manual of diagnosis and
therapy. 16th ed. Rahway, N.J.: Merck Research Laboratories, 1992.
2. Kennedy RS, Dunlap WP, Banderet LE, Smith MG, Houston CS. Cog-
nitive performance deficits in a simulated ascent climb of Mount Everest:
Operation Everest II. Aviat Space Environ Med 1989;60:99-104.
3. Litch JA. Subarachnoid hemorrhage at high altitude. West J Med 1997;
167:180-1.