Duration and Severity of Cerebral Hypoxia-Ischemia Determine
the Time-Course of ADC Changesand Histological Outcome
N. Miyasaka, *T. Nagaoka, j-T. Kuroiwa; *H. Akimoto, T. Haku, T. Kubota, T. Aso
Department of Obstetrics and Gynecology, *Neurosurgery, tNeuropathology,
School of Medicine, Tokyo Medical and Dental University, Tokyo, Japan
Introduction Cerebral hypoxia-&hernia (HI) is a major the pattern of ADC reduction (Fig. 2). The AR2* peak ratios
causeof perinatal brain damage. Recent studies with showed a threshold relationship for the time-course of ADC
diffusion-weighted MRIs (DWIs) have shown a biphasic decline evolution for eachof the HIic durations. In rats subjected to 15
of apparentdiffusion coefficient of brain water (ADC) following min HI, a AR2* peak ratio 1 50% was associatedwith transient
cerebralHI, i.e. primary ADC reduction during the acuteinsult, ADC reduction, and a AR2.* peak ratio 30% was associatedwith biphasic ADC reduction, and a
death. However, the circumstancesunder which such brain AR2* peak ratio < 30% was associatedwith permanentADC
damagetakes place have not been extensively examined. We reduction. The histology was closely associatedwith the
hypothesized that the duration of hypoxia and severity of pattern of ADC reduction. Selective neuronal necrosis was
ischemia defined the time-course and histological outcome of found in the ipsilateral parietal cortex ofthe rats which showed the
cerebralHI. To test this hypothesis, we measuredcerebral transient ADC reduction pattern, and pan-necrosiswas found in
perfusion and ADC using MRI techniques in young rats rats which showed the biphasic and the permanentADC
subjected to different durations of cerebralHI. The rats were reduction pattern.
monitored for up to 48 hours and histologically examined.
These MRI measurementsand histological findings were
correlated, and the factors which determine the outcome of
cerebralHI were investigated.
Materials and Methods Three-week-old male rats were
initially anesthetizedwith 5% isoflurane and then maintained
with 1.5% isoflurane in an air/oxygen mixture throughout the
experiment. After the left common carotid artery (CCA)
ligation and femoralvenous catheterinsertion were accomplished,
the rats were positioned in the magnet and normothermia was
maintained with a water blanket.
Images were acquiredin a 4.7 Tesla experimental system (Unity
INOVA, Varian, Palo Alto, Ca, USA) with a 33-cm horizontal B HI R (30m) R (60m) R (48h)
bore magnet, and an S-cm-ID quadrature coil was used for Fig. 1 Seriesof ADC mapsat the indicated time-points (B: before
radiofrequencyexcitation and reception. T2-weighted imaging hypoxia, HI: at the end ofthe insult, R(30m),R(60m),R(48h): 30
(T2WI) and DWI were performedwith conventional multislice min, 60 min, 48 hours afterresuscitat&kZgespectively).
spin-echo sequences (TlUTE msec= 1500/&O,with one average, Lw.2” peaklatin Individual AR2* peakratios at
2-mm slice thickness, 64x64 pixel matrix, 3 cm FOV), and two 80 the end of the eachdurations of
b-values (0, 1200 sec/mm2)were used for diffusion-weighting 70 i
the HIic insults in combination
along the x-, y-, and z- axesin turn. Perfusion-weightedimages 60
with the pattern of ADC
(PWIs) were obtained by dynamic susceptibility contrast- reduction.
50 0: The rats which showed
enhancedMRI (fast low-angle shot, TR/TE msec = 1518,with 8
n 2 “f
40 2 transient ADC reduction
one average,flip angle = 10, 2-mm slice thickness, 64x64 pixel 3% A: The rats which showed
matrix, 3 cm FOV). Eighty consecutive images of a slice (-2 30 P biphasic ADC reduction
mm from the bregma) were obtained, during which a bolus of 7.0 . i 0 : The rats which showed
P a
gadopentatedime&mine (Schering, 0.2 mmol/kg) was injected 10 permanent ADC reduction
manually through the femoral venous catheter. 0j
After baseline images were obtained, hypoxia was induced by 15mhHI 30mfnHI 6OminHI
lowering the inspired oxygen concentration to 8% by N2 dilution. Discussion and Conclusion Our study clearly demonstrated
T2WI and DWI were repeatedduring hypoxia. The rats were that the time-course of ADC changescould adopt three patterns,
randomly assignedto threegroups basedon different durations of namely transient, biphasic and permanentADC reduction
ADC reduction, (from the onset of ADC reduction to following the HIic insult, and that both the duration of hypoxia
resuscitation). The animals were resuscitated 15 min (n=l3), and severity of ischemia during the acute insult determined the
30 min (n=12) and 60 min (n=lO) after the onset of ADC pattern of ADC reduction. Such a two-dimensional threshold
reduction by increasing the oxygen concentration to 30%. concept (transverseaxis for the duration of hypoxia, and
PWIs were obtained just beforereoxygenation. All of the longitudinal axis for the severity of ischemia) is consistent with
animals were monitored for 60 min with MRI following that for stroke in which both the severity and duration of ischemia
reoxygenation. After awakening from anesthesia,the rats were were reported to define the development of cerebralinfarction3’.
returned to their cages,and T2WIs and DWls were obtained at 24 In addition, close associations between the transient ADC
and 48 hours afterHI. Finally, the animals were histologically reduction pattern and selective neuronal necrosis, and betweenthe
examined with HE staining. biphasic and permanentADC reduction patternsand pan-necrosis
Analyses of MR data were performedon a Sun Spark 10 were demonstratedon histology. This finding indicates that
workstation. ADC maps were calculated on a pixel-by-pixel ADC normalization upon resuscitation is not necessarily
basis using a standard equation described by Le Bihan et al”. associatedwith neuronal salvage. The application of theseMRI
The regions of interest (ROI) were drawn in both lateral parietal techniques in human and experimental cerebralHI may be useful
cortexes. The ROI was copied onto the PWIs, the signal in the evaluation of cerebralHI and for the establishment of
intensities were plotted over time, and the signal intensity data cerebroprotectivetherapies to prevent selective neuronal necrosis
were then converted to AR2*(t) using a standard technique. and delayed brain damage.
AR2* peak ratio was used as an index of cerebralperfusion. References
Results Three patterns of ADC changeswere found, namely 1) Dijkhuizen RM et al., Stroke, 29,695,1998.
the transient ADC reduction pattern (Fig. 1-a), the biphasic ADC 2) Le Bihan D et al., Radio&y, 168,497,1988.
reduction pattern (Fig. 1-b), and the permanentADC reduction 3) JonesTH et al., JNeurosurg. 54,773,1981.
pattern (Fig.l-c). Individual AR2* peak ratio at the end of the
eachdurations ofthe HIic insult were plotted in combination with