Correspondence of CT perfusion imaging to pathological manifestations in rabbit models of hyperacute cerebral infarction

摘要

BACKGROUND: Could the infarction be diagnosed quickly and accurately at the acute stage by CT perfusion imaging (CTPI) technology? Whether the images of CTPI will correspond with the pathological changes or not? All the questions need to be solved by experimental and clinical studies. OBJECTIVE: To reveal the rules of perfusion map changes and guide the early diagnosis of hyperacute cerebral infarction by analyzing the correlation of CTPI with pathological manifestations for hyperacute cerebral infarction. DESIGN: A randomized controlled animal experiment. SETTING: Experimental Center of Medical Radiology, Longgang Central Hospital of Shenzhen City. MATERIALS: Forty-two adult New Zealand rabbits of (2.6±0.5) kg, either male or female, were randomly divided into experimental group (n =36) and control group (n =6). Six rabbits in the experimental group were observed after ischemia for 0.5, 1, 2, 3, 4 and 6 hours respectively, and 1 rabbit in the control group was observed at each corresponding time point. METHODS: The experiments were carried out in the Experimental Center of Medical Radiology, Longgang Central Hospital of Shenzhen City from March 2003 to July 2004. Rabbit models of cerebral infarction were established by modified O'Brein method. ① The rabbits in the experimental group were scanned at 0.5, 1, 2, 3, 4 and 6 hours after ischemia respectively. The dynamic CT scan slice was 13 mm from the anterior edge of the frontal cortex, and six fake color functional images were obtained, including cerebral blood flow map (CBF map), cerebral blood volume map (CBV map), peak to enhancement map (PE map), flow without vessels map, time to peak map (TP map), time to start map (TS map). The manifestations and changes of the functional maps in different interval were observed. ② Bilateral symmetric ranges of interest (ROI) were drawn separately on the CBF map, CBV map, TP map and TS map. The blood flow parameters of focal and contralateral cerebral tissues could be obtained to calculate relative cerebral blood flow (rCBF, rCBF=focal CBF/contralateral CBF), relative cerebral blood volume (rCBV, rCBV= focal CBV/ contralateral CBV), a relative time to peak (rTP, rTP= focal TP–contralateral TP), a relative time to start (rTS, rTS= focal TP–contralateral TP). ③ The perfusion maps were input into AutoCAD software. The percents of ischemic cores and peri-ischemic areas accounting for contralateral cerebral hemisphere were calculated. ④ The animals were anesthetized and killed, then the cerebellum and low brain stem were taken out. The brain tissues were cut on coronal plane at 14 mm from the anterior edge of the frontal cortex, a 2-mm piece anterior to the incision, and a 3-mm piece posterior to the incision. The anterior piece was fixed, stained and observed. A 1-mm slice was cut from the front of the posterior piece tissues as electron microscope sample, the remnant was fixed and then taken out, and the location and size of stained "white" areas were observed as the reference for electron microscope sample. ⑤ The correlation between CTPI and pathological manifestations was observed. MAIN OUTCOME MEASURES: ① Laws of time and spatial changes of ischemic areas; ② Pathological changes of the ischemic tissues; ③ Correspondency between CTPI and pathological manifestations. RESULTS: ① Laws of time and spatial changes of ischemic areas: Relative ischemic-core areas were consistent in each perfusion map, increased incessantly along with the ischemic times. Relative peri-ischemic areas were inconsistent in each perfusion map, on CBF map from 1 to 6 hours after ischemia, the area of ischemic core increased from (1.503±0.523)% to (7.125±1.054)%, the ascending trend occurred. But the peri-ischemic areas showed a descending trend on CBF map, the areas decreased from (8.960±0.719)% to(5.445±0.884)% from 0.5 to 6 hours; The relative areas were the largest one on TP maps, the average value was (32.796±3.029)% at 0.5 hour after ischemia happening (60.540±1.683)% at 6 hours. The trend of ischemic areas was increased. No obvious change was observed on TS maps. ② Pathological changes of the ischemic tissues: Under light microscope, there was no obvious change at 0.5–2 hours after ischemia, edema at 3 hours, karyopycnosis at 4 hours and eosinophilous changes at 6 hours; Under electron microscope, there was edema in ischemic cores within 4 hours after ischemia, whereas karyopycnosis or structure vanished after 4 hours; Edema was observed in peri-ischemic areas. ③ Correlation between CTPI and pathological manifestations: On CTPI maps, the ischemic core was blue on CBF and CBV maps, black on TP and TS maps. Along with the ischemic times, the rCBF and rCBV decreased, whereas the rTP and rTS prolonged. Hemodynamic parameters were not significantly different within 2 hours of ischemia and 2 hours after ischemia. The rTP and rTS became 0 after 1 and 2 hours respectively. On CTPI maps the peri-ischemic area was red on CBF and CBV maps, red and yellow on TS maps, red on TP maps. Along with the ischemic times, the rCBF decreased, and the lowest level was always at about 20%, whereas the rTP and rTS prolonged. CONCLUSION: ① CTPI manifestations corresponded well with pathological findings, and it is a sensitive, stable and reliable technique to diagnose hyperacute cerebral infarction. ② TP map was more sensitive than CBF map and TS map in exhibiting the peri-ischemic areas, thus TP maps could be a good choice for observing peri-ischemic areas.