вход по аккаунту


Delayed OEF elevation is an unreliable index of tissue at risk.

код для вставкиСкачать
Delayed OEF Elevation is an Unreliable Index of
Tissue at Risk
Myron D. Ginsberg, MD
In their recent article, Furlan and colleagues [l], in an effort
to analyze “the fate of the ischemic penumbra,” present both
positron emission tomographic (PET) and late computed tomographic (CT) infarct data in 11 patients with ischemic
stroke who exhibited an elevated oxygen extraction fraction
(OEF) on PET scans. Of particular note is that all of these
patients were studied with a delay of 7hours or more following the onset of their ischemic ictus (mean ? SD, 10.5 ?
3.4 hours; range, 7-16 hours). As this range lies well beyond
the 3- to 4-hour time frame within which irreversible deterioration of the ischemic focus has been documented to occur in numerous experimental studies of focal ischemia in
rats [2-41, cats [5], and primates [6], a closer examination of
the data of Furlan and colleagues [I] might yield additional
insights as to the significance of elevated OEF when present
at 7 to 16 hours after stroke.
If the hypothesis were correct that elevated OEF at 7 to
16 hours faithfully represents “misery-perfused tissue still at
risk of ischemic infdrction, one would expect there to be a
significant correlation between the volume of the high-OEF
focus and the size of the eventual infarct on CT scan. In
Figure 1, we have plotted the total volume indices of the
high-OEF voxels-of-interest (VOIs) versus the volume indices of CT infarcts in the 11 patients of Furlan and colleagues
(data taken from their Table 1 [I]). As is evident from Figure 1, there is no suggestion of a correlation (R2= 0.05,
p = NS). From this analysis, we would conclude that by 7 to
16 hours after stroke, high-OEF zones per se no longer are
predictive of brain tissue still at risk of infdrction.
As Furlan and colleagues [ l ] further note, the high-OEF
VOIs in their patients could be assigned to one of four
groups, based upon 7- to 16-hour PET-cerebral blood flow
(PET-CBF) and chronic (approximately 50-day) CT criteria,
namely, type 1 = infarcted tissue in which earlier PET-CBF
was in the penumbral range (10-20 m1/100 gmimin); type
2 = noninfarcted tissue in which earlier PET-CBF was in
the penumbral range; type 3 = infarcted tissue in which earlier CBF was at “ischemic-core’’ levels of 0 to 10 m1/100
gm/min; and finally, type 4 = noninfarcted tissue in which
earlier PET-CBF was in the oligemic (ie, supraischemic)
range (above 20 mll100 gmlmin). W e carried out extensive
linear regression analyses to search for possible correlations
between the volumes of any of these high-OEF tissue types
(singly or in combination) and CT infarct volumes in the 11
patients, but in no instance could a significant correlation be
identified (range of R2,0.01-0.29; p = NS). We also
probed these data by nonparametric statistics. By the Spearman rank-correlation test, one significant correlation
emerged (shown in Fig 2), namely, the relative volume of
t-ype 4 (ie, oligemic noninfarcted) tissue on CBF-PET correlated inversely with relative CT infarct size (Spearman p =
-0.68, p = 0.02). That is, the larger the volume of highOEF tissue having CBF in the supraischemic range at 7 to
16 hours, the smaller the patient’s eventual infarct on CT
scan. This observation is viewed as merely confirming the
well-supported concept in the literature that, by 7 to 16
I .
R2= 0.05
5 t
Total High-OEF Volume Index (ml)
Fig 1. Lack of statistical correlation between total high-OEF
volume index and eventual CT infarct volume index in the
I I patients of Furlan et a1 [l].
Spearman R = 0.68
p = 0.02
Rank. High-OEF Zone 4
Fig 2. A n inverse linear correlation exists between the ordinal rank of high-OEF zone 4 (ie., oligemic noninfarcted tissue) and CT infarct volume index in tbe I I patients of
Furlun et al [l].
hours, brain tissue destined to infarct has CBF below 20 mll
100 gm/min.
The data of Furlan and colleagues [ l ] also allow one to
assess the significance of high-OEF VOIs observed at 7 to 16
hours in the larger context. O f their 26 consecutive patients
studied by PET, 6 were excluded because they evidently died
prior to late CT scanning, and 9 were excluded because of
“extensive hyperperfusion” without zones of elevated OEF.
Thus, of 20 chronically surviving patients, only 55% showed
any elevation of OEF when examined by PET at 6 hours or
more after stroke. When this entire patient base (n = 20) is
considered, the fraction of type 1 high-OEF tissue (ie, tissue
with CBF in the penumbral range at 7 to 16 hours that will
go on to infarction) contributes only 10 ? 4% (mean 5
SEM) to the ultimate CT infarct. This contrasts dramatically
with the relative volume of the early (1-2 hour) ischemic
penumbra, which has been shown in experimental studies to
constitute fully one-hulfof the entire ischemic focus [7,81.
Another reason to question the pathogenetic significance
of the “penumbra” as defined by OEF and CBF at 7 to 16
Copyright 0 1997 by the American Neurological Association
hours after stroke is readily apparent from Table 3 of Furlan
and colleagues [ I ] , which reveals that the high-OEF voxels of
type 1 (“penumbra” destined to infarct) and the high-OEF
voxels of type 2 (“penumbra” destined not to infarct) have
statistically indistinguishable, indeed virtually identical, values
for every measured PET variable, including CBF, cerebral
metabolic rate for oxygen (CMRo,), cerebral blood volume
(CBV), the CBF/CBV ratio, and OEF. Thus, by 7 to 16
hours after stroke, these conventional PET descriptors have
lost their ability to predict infarction.
Taken together, the considerations emerging from this
meta-analysis of the data of Furlan and colleagues [1] offer
strong support to the view that apparent elevations of OEF
observed afier 6 hours no longer denote tissue still at risk of irreversible injuty. These results contrast with the firm evidence
obtained from animal studies that earb metabolism-greaterthan-blood-flow dissociations observed within the first 1 to 2
hours reliably define tissue at risk of infarction [7-lo].
Stated differently, the data of Furlan and colleagues [l] appear to refute the concept that a significant viable ischemic
penumbra still exists after 6 hours, which might contribute
to ischemic infarction.
Department of Neurology, University o f Miami School of
Medicine, Miami, FL 33101
1. Furlan M, Marchal G, Viader F, et al. Spontaneous neurological recovery after stroke and the fare of the ischemic penumbra.
Ann Neurol 1996;40:216-226
2. Memezawa H , Smith M-I., Siesjo BK. Penumbral tissues salvaged by reprrfusion following middle cerebral artery occlusion
in rats. Srroke 1992;23:552-559
3. Kaplan B, Brint S, Tanabe J, ec al. Temporal rhresholds for
neocortical infarction in rats subjected to reversible focal cerebral ischemia. Stroke 1991:22:1032-1039
4. Buchan AM, Xue D , Slivka A. A new model of temporary focal
neocortical ischemia i n the rat. Stroke 1992;23:273-279
5. Heiss W-D, Rosner G. Functional recoveiy of cortical neurons
as related ro degree and duration of ischemia. Ann Nrurol
6. Jones T H , Morawetz KB, Crowell RM, et al. Thresholds of
focal cerebral ischemia in awake monkeys. J Neurosurg 1981;
7. Back T, Zhao W, Ginsberg MD. Three-dimensional imageanalysis of brain glucose inetabolism/blood flow uncoupling
and its electrophysiological correlates in the acute ischemic penumbra following middle cerebral artery occlusion. J Cereb
Blood Flow Metab 1995;15:566-577
8. Ginsberg M D , Belayev L, Zhao W, er al. The pathophysiological significance of nietabolism/blood flow uncoupling in acute
ischemic and traumatic brain injury. In: Krieglsrein J, ed. Pharmacology of cerebral ischemia-1996. Sturtgart: Medpharm Scientific Publishers (In press)
9. Back T, Ginsberg MD, Dietrich WD, Watson BD. Induction
of spreading depression in the ischemic hemisphere following
experimental middle cerebral artery occlusion: effect on infarct
morphology. J Cereb Blood Flow Merab 1996;16:202-213
10. Ginsberg M D , Pulsinelli WA. Edirorial: The ischemic penumbra, injury thresholds, and rhe rherapeucic window for acute
stroke. Ann Neurol 1994;36:553-554
Annals of Neurology
Vol 41
No 3
March 1997
J. C. Baron, M D , M. Furlan, MD,
and G. Marchal, MD, PhD
We are pleased that our clinical study [ I ] stimulated interest
from one of the world authorities in animal stroke research.
In patients with middle cerebral artery (MCA)-territory
stroke, we obtained (1) maps of cerebral blood flow (CBF),
oxygen extraction fraction (OEF), and cerebral nietabolic
rate oxygen (CMRo,) by means of positron emission tomography (PET), 7 to I 6 hours after onset: (2) scores of neurological deficit over 2 months: and (3) infarct mapping by
chronic-stage computed tomography (CT). Voxel-based analysis allowed us to identify in a subset of these patients tissue
fulfilling CBF and OEF criteria for penumbra, and to show
that a variable fraction of that tissue went on to infarction
while the other survived and subtended neurological recovery. This suggested that penumbral tissue with uncertain
outcome still existed 16 hours after onset in some patients.
D r Ginsberg concludes that our data “refute the concept
that a significant viable ischemic penumbra still exists after 6
hours, which might contribute to ischemic infarction.” As
this is serious criticism, D r Ginsberg’s arguments need to be
examined in detail.
elevated OEF at 7 to 16 hours
H e first states that “if
faithfully represents . . . tissue still at risk of ischemic infarction, one would expect there to be a significant correlation
between the volume of the high-OEF focus and the size of
the eventual infarct on CT scan.” Based on Table 1 of our
article [I], Dr Ginsberg fails to find such a correlation and
concludes that “by 7 to 16 hours after stroke, high-OEF
zones per se no longer are predictive of brain tissue still at
risk of infarction.” Although it may be that the extent of
high OEF correlates with final infarction volume in the case
of permanent MCA occlusion (MCAo) (a hypothesis testable
in our permanent MCAo baboon PET model [2]), this
should not apply to the case of early arterial recanalization
because efficient reperfusion immediately results in a dramatic fall in the OEF, with this tissue eventually escaping
infarction [ 3 ] ; indeed, in 9 baboons subjected to 3 to 6
hours of temporary MCAo, we observe no significant correlation between extent of the high-OEF zone (assessed 1 hour
after occlusion) and final infarct size (Spearman’s r =
-0.128, p = 0.74) [3]. In other words, high OEF is not in
itself predictive of subsequent infarction, and that was also
one major finding of our study in humans [I]. In the latter,
we speculated that in some of our patients, spontaneous recanalization after the PET study saved part of the penumbra;
however, this may have occurred in other patients despite
persistent occlusion, through, eg, delayed development of
collaterals, fluctuations in systemic or intracranial pressure,
or changes in arterial plasma ionic composition or blood
gases. Thus, depending on events subsequent to PET, a correlation between high-OEF extent and infarct size may or
may not exist, so the lack of such a correlation in our study
[ 11 by no means disproves the hypothesis that at least part of
the high-OEF tissue was indeed at risk of infarction.
Three earlier clinical reports actually provide ample evidence that this is so much later than 4 to 6 hours after
stroke. In one [ 4 ] ,the high-OEF tissue observed up to 36
hours consistently progressed to severe metabolic deteriora-
Без категории
Размер файла
239 Кб
unreliable, oef, elevations, delayed, index, tissue, risk
Пожаловаться на содержимое документа