Infarct Volume as a Surrogate or Auxiliary Outcome

Measure in Ischemic Stroke Clinical Trials

Jeffrey L. Saver, MD; Karen C. Johnston, MD; Daniel Homer, MD; Robert Wityk, MD;
Walter Koroshetz, MD; Laura L. Truskowski, PhD; E. Clarke Haley, MD;
for the RANTTAS Investigators

Background and Purpose—Reduction in infarct volume is the standard measure of therapeutic success in animal stroke
models. Reduction in infarct volume has been advocated as a biological surrogate or auxiliary outcome measure for
human stroke clinical trials to replace or supplement deficit, disability, and global clinical scales. However, few studies
have investigated correlations between infarct volume and clinical end points in acute ischemic stroke patients.
Methods—CT scans at days 6 to 11 were acquired prospectively in 191 fully eligible patients enrolled in the Randomized
Trial of Tirilazad Mesylate in Patients With Acute Stroke (RANTTAS). Patients were enrolled within 6 hours of onset
of stroke in any vessel distribution. Infarct volume was measured by operator-assisted computerized planimetry.
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017

Results—One hundred thirty-two patients had visible new supratentorial infarcts, with median infarct volume of 28.0 cm3
(interquartile range, 9.0 to 93.0 cm3). Fifty-nine patients had no visible new infarct. Correlations with standard 3-month
outcome scales and mortality were as follows: Barthel Index, r50.43; Glasgow Outcome Scale, r50.53; National
Institutes of Health Stroke Scale, r50.54; mortality, r50.31. For visible infarcts alone, correlations were as follows: BI,
r50.46; GOS, r50.59; NIHSS, r50.56; mortality, r50.32.
Conclusions—Subacute CT infarct volume correlates moderately with 3-month clinical outcome as assessed by widely
used neurological and functional assessment scales. The modesty of this linkage constrains the use of infarct volume as
a surrogate end point in ischemic stroke clinical trials. (Stroke. 1999;30:293-298.)
Key Words: cerebral infarction n clinical trials n stroke assessment n stroke outcome n tomography, x-ray computed

R andomized clinical trials employ a variety of end points

to determine the effectiveness of a novel agent. For
definitive phase 3 trials, the primary end point should be a
clinical outcome of direct relevance to the patient.1 However,
important clinical outcomes are often infrequent and delayed,
and phase 3 trials consequently may be time-consuming and
expensive. In phase 2 trials, the goal is rapid, inexpensive
screening of new agents and dosages for biological activity
and safety. For these purposes, using clinical outcome mea-
sures alone may mandate impractical sample sizes. Clinical
trialists and statisticians therefore have sought intermediate,
surrogate end points.
A surrogate end point, to quote Temple’s definition from
among many available, “is a laboratory measurement or a
physical sign used as a substitute for a clinically meaningful
endpoint that measures directly how a patient feels, functions
or survives. Changes induced by a therapy on a surrogate
endpoint are expected to reflect changes in a clinically
meaningful endpoint.”2 The concept of a surrogate end point
has been precisely made operational with the use of statistical
formulas.3,4 Ideally, surrogate end points should be simple to
ascertain, easily quantifiable, inexpensive, and tightly corre-
lated with targeted clinical outcomes.5 A related concept is
that of an auxiliary end point. An auxiliary end point is an
intermediate biomarker for a true clinical end point that, by
demonstrating a similar response to an intervention, supple-
ments or strengthens standard analysis of true clinical end
point data.6
Phase 3 clinical trials of therapies for acute ischemic stroke
currently use as primary end points scores at 3 to 6 months on
rated measures of activities of daily living, disability, neuro-
logical deficit, and/or global outcome. Drawbacks of these
instruments include acceptable but not excellent interrater
reliability,7 insensitivity to fine variations in treatment effec-
tiveness, the costs of maintaining patient follow-up over a 3-
to 6-month period, and extreme variability. In contrast,

Received August 28, 1998; final revision received November 11, 1998; accepted November 11, 1998.
From the Stroke Center and Department of Neurology, University of California, Los Angeles (J.L.S.); the Department of Neurology, University of
Virginia (Charlottesville) (K.C.J., L.L.T., E.C.H.); the Department of Neurology, Northwestern Medical School, Evanston, Ill (D.H.); the Department of
Neurology, Johns Hopkins University, Baltimore, MD (R.W.); and the Department of Neurology, Harvard Medical School, Boston, Mass (W.K.).
The RANTTAS Investigators are listed in the Appendix of Reference 10.
Presented in part at the 121st Annual Meeting of the American Neurological Association, Miami, Fla, October 13–16, 1996.
Correspondence to Jeffrey L. Saver, MD, UCLA Stroke Center, Reed Neurologic Research Center, 710 Westwood Plaza, Los Angeles, CA 90095.
E-mail jsaver@ucla.edu
© 1999 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org

293
 294 Infarct Volume as a Stroke Trial Surrogate Outcome Measure
measurement of infarct volume is the standard gauge of
therapeutic efficacy in animal stroke models. Decrease in
infarct volume has been suggested as an appropriate surrogate
or auxiliary outcome measure for human stroke trials.8,9
However, few reports have analyzed correlations between
infarct volume and clinical end points in stroke trial
populations.
We studied correlations between infarct volume and loca-
tion and clinical outcomes in the Randomized Trial of
Tirilazad Mesylate in Patients With Acute Stroke
(RANTTAS).
Subjects and Methods
RANTTAS was a multicenter, randomized, double-blind, vehicle-
controlled trial to evaluate the safety and efficacy of intravenous
tirilazad mesylate in patients with acute ischemic stroke.10 Tirilazad
is a nonglucocorticoid 21-aminosteroid that inhibits lipid peroxida-
tion in vitro and reduces infarct volume in animal models of focal
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017
ischemia. Patients were enrolled within 6 hours of ischemic stroke in
any vessel distribution.
Cranial CT scans were performed on days 6 to 11 in a prospective
random sample of 50% of fully eligible subjects. Noncontrast scans
were obtained according to a standardized protocol that specified
5-mm slice intervals. Volume of cerebral infarction was measured
centrally by an investigator blinded to treatment group using com-
puterized planimetric techniques.8 Localization analysis was per-
formed centrally by investigators blinded to treatment group, using a
checklist of anatomic structures encompassed by the lesion. Involve-
ment or sparing was noted for the following 9 regions: frontal,
parietal, temporal, occipital, corona radiata/internal capsule, basal
ganglia, thalamus, cerebellum, and brain stem.
Patients were included in this study if they fulfilled the following
criteria: technically adequate CT scans were performed in the day 6
to 11 window, with the films received at the coordinating center;
infarct location was determined to be hemispheric; no prestroke
disability was present (determined by history obtained at baseline
entry into the trial); and 3-month outcome evaluations were obtained
between days 76 and 106 after entry, or patients died before that
window. Patients with brain stem/cerebellar strokes were excluded
from this analysis because infarcts in this location are not reliably
detected by CT due to bone artifacts. Patients with prestroke
disability, from either prior stroke or other causes, were excluded
because of the potential confounding effects on the outcome disabil-
ity measures. Patients who had their outcome evaluations performed
outside the day 76 to 106 window were excluded from this study
because of the potential confounding effects of changing neurolog-
ical and functional status in patients recovering from stroke.
Testing for a treatment effect with the use of the Kruskal-Wallis
test demonstrated no significant effect of treatment assignment on
infarct volume. The tirilazad group median infarct volume was 11.1
cm3 (interquartile [IQ] range, 0 to 57.8 cm3 ) versus vehicle median
infarct volume of 8.5 cm3 (IQ range, 0 to 84.7 cm3) (P50.78,
Kruskal-Wallis test). Accordingly, both groups were combined for
further analyses of relations between infarct volume and outcome.
Clinical outcome measures performed from days 76 to 106
included the modified National Institutes of Health Stroke Scale
(NIHSS),11 a measure of neurological deficit; the Barthel Activities
of Daily Living Index (BI)12; and the Glasgow Outcome Scale
(GOS),13 a global rating of disability. Patients who died before their
3-month outcome evaluation were excluded from NIHSS and BI
analyses because inclusion would have required imputation of a
score for death. For this reason, mortality was examined separately.
Since the GOS includes a rating for death, dead patients were
included in the GOS analysis. In addition to examining correlations
of infarct volume with the full scales, the predictive value of infarct
volume in determining a favorable outcome was examined. Criteria
for favorable outcomes predefined by the protocol were BI score of
60 to 100 and GOS score #2 (good recovery or moderate disability).
TABLE 1. Correlations of CT Infarct Volume and 3-Month
Functional Outcome Measures (Correlation Coefficients)
Outcome Measure All Infarcts (95% CI) Visible Infarcts Only (95% CI)
BI (n5150) 0.43 (0.28–0.58) 0.46 (0.29–0.64)
GOS (n5170) 0.53 (0.40–0.65) 0.59 (0.47–0.71)
NIHSS (n5131) 0.54 (0.41–0.68) 0.56 (0.40–0.71)
Mortality (n5191) 0.31 (0.18–0.45) 0.32 (0.16–0.49)
Additionally, for this analysis, a favorable outcome on the NIHSS
was defined as a score #1.14
In statistical analysis, Spearman’s r15 was used to compute the
correlation between infarct volume and each of the 3-month clinical
outcome measures (BI, GOS, and NIHSS), as well as mortality.
Logistic regression was used to calculate odds ratios for favorable
outcome for each cubic centimeter increase in infarct volume.16 The
influence of infarct location, lateralization, and presence or absence
of cortical involvement was also examined with either Spearman’s r
or multivariable logistic regression.
Results
One hundred ninety-one patients met the inclusion criteria for
this study. Of 256 CT studies obtained, 3 were technically
inadequate, 17 had clinical brain stem/cerebellar infarcts, 25
were performed outside the day 6 to 11 window, and 20 had
prestroke disability. Of the 191, 20 patients died during
follow-up and were therefore excluded from the BI and
NIHSS analyses. Twenty-one patients had their 3-month
follow-up scores performed outside the day 76 to 106
window and were therefore excluded from the BI, GOS, and
NIHSS analyses. An additional 17 patients did not have the
NIHSS performed at follow-up, and 2 more had 1 or more
missing subscores at follow-up. The total number of subjects
for the mortality analyses was therefore 191; for the GOS,
n5170; for the BI, n5150; and for the NIHSS, n5131.
Among the 191 study patients, 57% were male, 43% were
female, and the mean6SD age was 69612 years. The
RANTTAS qualifying stroke was the first-ever stroke in
64%, while 15% had prior clinical strokes, 16% had prior
silent strokes (evident on baseline CT only), and prior history
information was unavailable in 1%. Among those with a
clinical history of prior stroke, 61% had an old stroke visible
on baseline CT. The median NIHSS score on entry was 10
(IQ range, 5 to 16).
On the day 6 to 11 CT, 132 patients had visible new
infarcts and 59 had no visible infarcts. The overall median
infarct volume was 10.5 cm3 (IQ range, 0 to 65.0 cm3). For
the 132 visible infarcts, the median infarct volume was 28.0
cm3 (IQ range, 9.0 to 93.0 cm3).
The correlations between infarct volume and 3-month
outcome are shown in Table 1. Correlation coefficients
ranged from 0.31 for mortality to 0.54 for the NIHSS.
Limiting the analysis to just the 132 patients with new visible
infarcts did not substantially change the results. Imputing
worst outcome scores on the BI and the NIHSS for patients
who died before 3-month follow-up also did not significantly
alter correlation values (data not shown). Figures 1 to 4
graphically illustrate the relationships between infarct volume
and clinical outcome.

Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017
Analyses of infarct lateralization and pure subcortical, pure
cortical, and mixed subcortical-cortical subgroups failed to
yield significant improvements in the correlations (data not
shown). Analyses of infarct localization by neuroanatomic
region were not sufficiently powered to detect major local-
ization effects (1 data table filed with the National Auxiliary
Publications Service).
Table 2 shows the odds ratios and 95% CIs for a dichot-
omized favorable outcome associated with a unit increase in
infarct volume. None of the CIs include 1.0. Table 3 shows
the median infarct volumes for favorable and unfavorable
outcomes among the different outcome measures. Patients
with favorable outcomes had median infarct volumes ,10
cm3, while patients with unfavorable outcomes (except for the
NIHSS, which scored anything other than complete or nearly
complete recovery on the neurological examination as unfa-
vorable) had median infarct volumes .100 cm3.
Discussion
In this sample from a large, multicenter clinical trial, infarct
volume correlated moderately with standard clinical mea-
Saver et al February 1999 295
Figure 1. Scatterplot of infarct volume vs 3-month
NIHSS.
sures of stroke outcome. Correlation coefficients were in the
0.43 to 0.54 range for infarct volume alone versus 3 common
clinical outcome measures after stroke: the BI, NIHSS,
and GOS.
Comparable prior studies are sparse and contradictory.
Several small series demonstrated no significant statistical
correlation between CT lesion volume and clinical measures
(BI, NIHSS, Rankin Disability Scale).9,17,18 Other, generally
larger, series did show correlations between subacute CT
lesion size and clinical scales (NIHSS, Rankin Disability
Scale, Oxford Disability Scale, aphasia severity scale).7,19 –21
Compared with these earlier investigations, the RANTTAS
data set is unique in size, geographic diversity, and actual
participation in an acute stroke clinical trial. Our findings
consequently are more likely to approximate the true rela-
tionship between subacute CT infarct volumes and 3-month
clinical measures, especially in clinical trial settings.
Are correlation coefficients in the 0.31 to 0.54 range
adequate to endorse CT infarct volume as a potential surro-
gate end point for clinical measures of outcome? This linkage
Figure 2. Scatterplot of infarct volume vs
3-month BI.
 296 Infarct Volume as a Stroke Trial Surrogate Outcome Measure
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017
clearly falls short of an ideal biomarker and is insufficient to
allow CT infarct volume to serve as a replacement end point
for clinical markers in pivotal phase 3 efficacy trials, for
which false-positive and false-negative rates must be low,
usually 2.5% to 10%.1 However, precedent suggests that this
degree of linkage may be sufficient to permit adopting CT
infarct size as an auxiliary measure of outcome in assessing
phase 3 trials. The most influential auxiliary outcome mea-
sure in neurological pharmaceutical trials in the 1990s has
been MRI-measured plaque volume in multiple sclerosis.22
However, the correlation between MRI plaque volume and
the expanded Kurtzke Disability Scale is only 0.23 to 0.33,23
substantially less than the correlations we observed between
CT infarct volume and stroke disability and activity of daily
living scales.
An important factor potentially attenuating the relation
between infarct size and clinical outcome measures is
lesion location. The site of injury is a consequential
additional variable determining the extent of functional
deficits from stroke. The effect of location reflects human
Figure 3. Box and whiskers plot of Infarct volume
vs GOS. GR indicates good recovery; MD, moder-
ate disability; and SD, severe disability. The hori-
zontal line indicates the median value, the box
delineates the middle 50% of values (the IQ range),
the whiskers demarcate values within 1.5 IQ
ranges, and the small box with plus symbol indi-
cates observations .3 IQ ranges below the first
quartile or above the third quartile.
brain organization into regionally distributed, large-
scale neural networks subserving specialized functions,
including motor control and language, with disparate
impact on clinical outcome. In the RANTTAS data set,
limited cell size in individual neuroanatomic regions
limited exploration of and adjustment for the effects of
lesion localization.
Several additional factors besides lesion location likely
weaken correlations between infarct volume and clinical
outcome. Confounding variables include the following:
(1) patient age; (2) patient sex; (3) baseline cerebral atrophy;
(4) prior symptomatic and silent infarcts; (5) prior nonvascu-
lar cerebral insults; (6) “fogging” effect on 1-week CT;
(7) variable development of medical complications of stroke,
including pneumonia, deep venous thrombosis, and pulmo-
nary emboli; (8) interindividual variations in neuroplasticity
and recovery capacity; (9) variable exposure to pharmacolog-
ical agents that secondarily enhance or inhibit neuroplasticity
and recovery; and (10) degree of family and community
social supports.24,25
Figure 4. Box and whiskers plot of infarct volume
vs mortality. Marking conventions are as detailed
for Figure 3.

TABLE 2. Influence of Infarct Volume on Favorable BI, GOS,
and NIHSS Outcomes and on Mortality
Outcome Measure Odds Ratio (95% CI)*
BI 1.019 (1.010–1.028)
GOS 1.026 (1.016–1.035)
NIHSS 1.050 (1.028–1.073)
Mortality 0.988 (0.982–0.993)
3
*Odds ratios are for each 1 mm decrease in infarct volume and are adjusted
for age, sex, prior stroke, and old infarct on CT scan.
In our population, day 6 to 11 CT failed to visualize an
infarction in 33% of patients. This detection failure rate is
consistent with those as high as 23% to 59% observed in prior
studies.7,26 –29 In part, this reflects intrinsic limitations in the
CT technique, including limitations on spatial resolution,
fogging effect of luxury perfusion obscuring hypodensity,
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017
and skull x-ray beam hardening.30 An additional factor in our
population is the effect of hyperacute patient selection. A
proportion of patients enrolled within the first 6 hours of
ischemic symptom onset may show early subsequent im-
provement and clinically exhibit no or minimal residual
deficit at follow-up. These patients are likely to have smaller
volume infarcts than patients with significant deficits at 24
hours. Correlation coefficient values between infarct volumes
and clinical outcomes were not substantially different when
analysis was confined only to those who had a visible infarct,
compared with all eligible patients. Thus, including the cases
with no visible infarct did not significantly alter our correla-
tion results. However, the lack of visible CT lesions in a
substantial minority of patients, a “floor effect,” is an impor-
tant constraint on the use of CT measures of lesion volume as
outcome markers.
Two other potential limitations of our study may be noted.
We obtained CT scans over a relatively broad day 6 to 11
window. Mass effect from cytotoxic edema may be variably
resolving during this period and may be expected to be more
substantial in earlier than later scans, a possible confounding
effect. To address this issue, we performed separate correla-
tion analyses for early CT window (days 6 to 8) and late CT
window (days 9 to 11) patients (data not shown). There was
mild improvement in the ability of CT to predict mortality,
but there were no major differences for BI, GOS, and NIHSS
correlations. Second, approximately 10% to 15% of patients
TABLE 3. Median Infarct Volumes in Patients With Favorable
and Unfavorable Outcomes
Outcome Measure Response Infarct Volume, cm3
BI Favorable outcome 3.5 (0–25.7)
Alive with unfavorable outcome 100.2 (25.1–156.9)
GOS Favorable outcome 3.2 (0–21.7)
Unfavorable outcome 108.0 (25.1–181.4)
NIHSS Favorable outcome 0.7 (0–67.6)
Alive with unfavorable outcome 46.4 (1.9–108.7)
Mortality Surviving 7.4 (0–49.2)
Dead 123.9 (19.4–213.4)
Values for infarct volume are median (IQ range).
Saver et al February 1999 297
receiving CT scans had day 76 to 106 data missing, their final
trial visit having occurred before or after this window. To
analyze whether including these patients would alter our
findings, we performed a last observation carried forward
analysis for the BI, GOS, and NIHSS (data not shown).
Correlation coefficients with infarct volume did not materi-
ally change. Thus, neither the use of the day 6 to 11 CT
window nor the day 76 to 106 clinical outcome window
substantially lessened correlations between CT infarct vol-
ume and clinical outcome.
The modesty of the correlation between CT infarct volume
and standard measures of clinical outcome suggests that CT
volume could be considered a somewhat useful but imperfect
auxiliary outcome marker for ischemic stroke studies. New
imaging techniques may provide better early biological mark-
ers of treatment effectiveness. For example, T2-weighted MR
measures have greater spatial resolution and sensitivity to
ischemic injury than CT attenuation. Few series have com-
pared conventional MR measures of infarct volume and
clinical end points, but the largest such study supports the
hypothesis that MR measures may more closely correlate
with functional outcome. Among 50 patients, MR scans
performed 2.6 to 44.8 weeks after onset were compared with
NIHSS scores obtained the same day and BI scores obtained
12 weeks after onset. T2-weighted infarct volume measures
correlated somewhat more closely with BI and NIHSS scores
than did CT infarct volumes in the RANTTAS data set.31
Even more promising as candidate auxiliary outcome mea-
sures are novel MRI diffusion and perfusion techniques that
allow early patient-specific volumetric assessment of the
region at ischemic risk and late determination of the volumes
of infarcted and salvaged tissues.32,33 These methods, how-
ever, will require rigorous evaluation before their acceptance
as adequate auxiliary or surrogate clinical trial outcome
measures.
It is important to note that imaging markers, no matter how
accurate, can never fully replace clinical assessments of
long-term outcome. Dissociated effects of treatments on
lesion volume and clinical outcome can occur through a
variety of mechanisms. For example, an effective acute
neuroprotective agent that had severe but delayed neurotoxic
effects would yield favorable early imaging outcomes but
unfavorable long-term clinical outcomes. Conversely, a
neuroplasticity-enhancing agent that did not alter initial
injury but improved functional recovery in uninjured regions
would yield unfavorable early imaging results but favorable
long-term clinical outcomes. Accordingly, clinical outcome
scales should continue to serve as the primary end points in
definitive, phase 3 randomized clinical trials of novel thera-
pies for acute ischemic stroke.
Acknowledgments
This work was supported by grants from the National Institutes of
Neurological Disorders and Stroke, National Institutes of Health
(RO1-NS31554), Pharmacia and Upjohn Company (Kalamazoo,
Mich), Pharmaceutical Research and Manufacturers (Dr Johnston),
and the University of Virginia (Dr Johnston). We thank Wayne
Alves for expert assistance with statistical graphing. Two additional
data tables are available for a nominal fee by writing ASIS/NAPS,
c/o Burrows Systems, 248 Hempstead Turnpike, West Hempstead,
 298 Infarct Volume as a Stroke Trial Surrogate Outcome Measure
NY 11552-2664. One explores the influence of involvement of 9
neuroanatomic regions on clinical outcome; the other lists sample
sizes required for clinical trials using different clinical and CT end
points, based on measure variability observed in the RANTTAS trial.
References
1. Fleming TR, DeMets DL. Surrogate end points in clinical trials: are we
being misled? Ann Intern Med. 1996;125:605– 613.
2. Temple RJ. A regulatory authority’s opinion about surrogate endpoints.
In: Nimmo WS, Tucker GT, eds. Clinical Measurement in Drug Eval-
uation. New York, NY: John Wiley & Sons Inc; 1995.
3. Prentice RL Surrogate endpoints in clinical trials: definition and opera-
tional criteria. Stat Med. 1989;8:431– 440.
4. Boissel J-P, Collet J-P, Moleur P, Haugh M. Surrogate endpoints: a basis
for a rational approach. Eur J Clin Pharmacol. 1992;43:235–244.
5. Brass LM. Trial designs issues: endpoints and sample size. Cerebrovasc
Dis. 1995;5(suppl 1):3–11.
6. Fleming TR, Prentice RL, Pepe MS, Glidden D. Surrogate and auxiliary
endpoints in clinical trials, with potential applications in cancer and AIDS
research. Stat Med. 1994;13:955–968.
7. Brott T, Adams HP, Olinger CP, Marler JR, Barsan WG, Biller J, Spilker
J, Holleran R, Eberle R, Hertzberg V, Rorick M, Moomaw CJ, Walker M.
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017
Developing measurements of acute cerebral infarction: a clinical exami-
nation scale. Stroke. 1989;20:864 – 870.
8. Brott T, Marler JR, Olinger CP, Adams HP, Tomsick T, Barsan WG,
Biller J, Eberle R, Hertzberg V, Walker M. Measurements of acute
cerebral infarction: lesion size by computed tomography. Stroke. 1989;
20:871– 875.
9. Chua MC, Davis SM, Infeld B, Rossiter SC, Tress BM, Hopper JL.
Prediction of functional outcome and tissue loss in acute cortical
infarction. Arch Neurol. 1995;52:496 –500.
10. The RANTTAS Investigators. A Randomized Trial of Tirilazad Mesylate
in Patients With Acute Stroke (RANTTAS). Stroke. 1996;27:1453–1458.
11. Lyden P, Brott T, Tilley B, Welch KM, Mascha EJ, Levine S, Haley EC,
Grotta J, Marler J. Improved reliability of the NIH Stroke Scale using
video training. Stroke. 1994;25:2220 –2226.
12. Mahoney F, Barthel D. Functional evaluation: the Barthel Index. Md Med
J. 1965;2:61– 65.
13. Jennett B, Bond M. Assessment of outcome after severe brain damage: a
practical scale. Lancet. 1975;1:480 – 484.
14. NINDS rt-PA Stroke Study Group. Tissue plasminogen activator for
acute ischemic stroke. N Engl J Med. 1995;333:1581–1587.
15. Conover WJ. Practical Nonparametric Statistics. New York, NY: John
Wiley & Sons Inc; 1980.
16. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY:
John Wiley & Sons Inc; 1989.
17. Lyden PD, Zweifler R, Mahdavi Z, Lonzo L. A rapid, reliable, and valid
method for measuring infarct and brain compartment volumes from
computed tomographic scans. Stroke. 1994;25:2421–2428.
18. Streifler JY, Beloosesky Y, Grinblat Y. Infarct size and functional
recovery do not always correlate: a brain CT study in acute stroke
patients. Neurology. 1995;45(suppl 4):A331. Abstract.
19. Olsen TS. Outcome following occlusion of the middle cerebral artery.
Acta Neurol Scand. 1991;83:254 –258.
20. Finocchi C, Gandolfo C, Gasparetto B, Del Sette M, Croce R, Loeb C.
Value of early variables as predictors of short-term outcome in patients
with acute focal cerebral ischemia. Ital J Neurol Sci. 1996;17:341–346.
21. Naeser MA, Hayward RW, Laughlin SA, Zata LM. Quantitative CT scan
studies in aphasia, I: infarct size and CT numbers. Brain Lang. 1981;12:
140 –164.
22. Paty DW, Li DK, UBC MS/MRI Study Group, IFB MS Study Group.
Interferon beta-1b is effective in relapsing-remitting multiple sclerosis, II:
MRI analysis results of a multicenter, randomized, double-blind, placebo-
controlled trial. Neurology. 1993;43:662– 667.
23. Miller DH, Albert PS, Barkhof F, Francis G, Frank JA, Hodgkinson S,
Lublin FD, Paty DW, Reingold SC, Simon J. Guidelines for the use of
magnetic resonance techniques in monitoring the treatment of multiple
sclerosis. Ann Neurol. 1996;39:6 –16.
24. Jongbloed L. Prediction of function after stroke: a critical review. Stroke.
1986;17:765–776.
25. Goldstein LB, for the Sygen in Acute Stroke Study Investigators.
Common drugs may influence motor recovery after stroke. Neurology.
1995;45:865– 871.
26. Chamorro A, Vila N, Ascaso C, Saiz A, Montalvo J, Alonso P, Tolosa E.
Early prediction of stroke severity: role of the erythrocyte sedimentation
rate. Stroke. 1995;26:573–576.
27. Lindgren A, Norvving B, Rudling O, Johansson BB. Comparison of
clinical and neuroradiological findings in first-ever stroke. Stroke. 1994;
25:1371–1377.
28. Bryan RN, Levy LM, Whitlow WD, Killian JM, Preziosi TJ, Rosario JA.
Diagnosis of acute cerebral infarction: comparison of CT and MR
imaging. AJNR Am J Neuroradiol. 1991;12:611– 620.
29. Tatemichi TK, Mohr JP, Rubinstein LV, Kase CS, Nichols FT, Price TR,
Wolf PA. CT findings and clinical course in acute stroke: the NINCDS
Pilot Stroke Data Bank. Stroke. 1985;16:138.
30. Skriver EB, Olsen TS. Transient disappearance of cerebral infarcts on CT
scan: the so-called fogging effect. Neuroradiology. 1981;22:61– 65.
31. Lovblad K-O, Baird AE, Schlaug G, Benfield A, Siewert B, Voetsch B,
Connor A, Burzynski C, Edelman R, Warach S. Ischemic lesion volumes
in acute stroke by diffusion-weighted magnetic resonance imaging cor-
relate with clinical outcome. Ann Neurol. 1997;42:164 –170.
32. Warach S, Dashe JF, Edelman RR. Clinical outcome in ischemic stroke
predicted by early diffusion-weighted and perfusion magnetic resonance
imaging: a preliminary analysis. J Cereb Blood Flow Metab. 1996;16:
53–59.
33. Warach S, Benfield A, Schlaug G, Siewert B, Edelman RR. Reduction of
lesion volume in human stroke by citicoline detected by diffusion
weighted magnetic resonance imaging: a pilot study. Ann Neurol. 1996;
40:527.
 Infarct Volume as a Surrogate or Auxiliary Outcome Measure in Ischemic Stroke Clinical
Trials
Jeffrey L. Saver, Karen C. Johnston, Daniel Homer, Robert Wityk, Walter Koroshetz, Laura L.
Truskowski and E. Clarke Haley
for the RANTTAS Investigators
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017

Stroke. 1999;30:293-298
doi: 10.1161/01.STR.30.2.293
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1999 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/30/2/293

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Stroke is online at:

http://stroke.ahajournals.org//subscriptions/