Cerebral blood flow thresholds predicting new hypoattenuation areas due to macrovascular ischemia during the acute phase of severe and complicated aneurysmal subarachnoid hemorrhage. A preliminary study.

BACKGROUND: Focal ischemia may affect patients with aneurysmal subarachnoid hemorrhage (SAH), and the potential evolution of cerebral infarction may greatly influence the patients' outcome. The aim of the study was to assess the values of regional cortical cerebral blood flow (rCBF) thresholds predictive for ischemia during the acute phase of SAH. METHODS: In 34 patients affected by poor grade or complicated SAH, 52 pairs of Xenon-CT (Xe-CT) studies of regional CBF were analyzed, in which the follow-up Xe-CT study was obtained no later than 72 hours after the baseline study. Corresponding cortical ROIs were placed in the perimeter of the cortex on both the Xe-CT studies. A blinded, experienced neuroradiologist classified for each ROI, the development of a new hypoattenuation at the unenhanced CT images included in the follow-up Xe-CT, while another independent investigator collected rCBF levels of the ROI in the baseline Xe-CT study. FINDINGS: New hypoattenuation developed in 3.94% of the ROIs in the paired follow-up Xe-CT studies, and these evolving ROIs were associated with a lower rCBF in baseline Xe-CT. However, the positive predictive value of rCBF levels for the development of new hypoattenuation was only moderately predictive (28.3%) for very low physiological values (5 ml/100gr/min). CONCLUSIONS: The results suggest that there is no absolute rCBF threshold ofischemia in severe and complicated SAH patients and that the rCBF values are only moderately predictive at levels lower than previously described.

Regional cerebral blood flow levels as measured by xenon-CT in vascular territorial low-density areas after subarachnoid hemorrhage are not always ischemic.

INTRODUCTION: The aim of this study was to assess regional cerebral blood flow (rCBV) in areas of CT hypoattenuation appearing in the postoperative period in patients treated for aneurysmal subarachnoid hemorrhage (SAH) using xenon-enhanced CT scanning (Xe-CT). METHODS: We analyzed 15 patients (5 male and 10 female; mean age 49.7+/-12.1 years) with SAH on CT performed on admission to hospital and who showed a low-density area within a well-defined vascular territory on CT scans after clipping or coiling of a saccular aneurysm. All zones of hypoattenuation were larger than 1 cm(2) and showed signs of a mass effect suggesting a subacute phase of evolution. Two aneurysms were detected in two patients. Aneurysms were located in the middle cerebral artery (n=7), in the anterior communicating artery (n=6), in the internal carotid artery (n=3), and in the posterior communicating artery (n=1). Treatments were surgical (n=8), endovascular (n=2) or both (n=1). A total of 36 Xe-CT studies were performed and rCBF values were measured in two different regions of interest (ROI): the low-density area, and an area of normal-appearing brain tissue located symmetrically in the contralateral hemisphere. RESULTS: rCBF levels were significantly lower in the low-density area than in the contralateral normal-appearing area (P<0.01). In the low-density areas, irreversible ischemia (CBF <10 ml/100 g per minute) was present in 11/36 lesions (30.6%), ischemic penumbra (CBF 10-20 ml/100 g per minute) and oligemia (CBF 20-34 ml/100 g per minute) in 8/36 lesions (22.2%), relative hyperemia (CBF 34-55 ml/100 g per minute) in 7/36 lesions (19.4%), and absolute hyperemia (CBF >55 ml/100 g per minute) in 2/36 lesions (5.6%). CONCLUSION: Our study confirmed that rCBF is reduced in new low-density lesions related to specific vascular territories. However, only about one-third of the lesions showed rCBF levels consistent with irreversible ischemia and in a relatively high proportion of lesions, rCBF levels indicated penumbral, oligemic and hyperemic areas.

Diagnostic impact of the spectrum of ischemic cerebral blood flow thresholds in sedated subarachnoid hemorrhage patients.

OBJECTIVE: Ischemia is the main cause of secondary damage in subarachnoid hemorrhage (SAH). Cerebral blood flow (CBF) measurement is useful to detect critical values. We analyzed the diagnostic impact of CBF ischemic thresholds to predict a new low attenuation area on computed tomography (CT) due to failure of large vessel perfusion. METHODS: We analyzed 48 xenon CT (Xe-CT) studies from 10 patients with SAH. CBF measurements were obtained by means of Xe-CT and cortical regions of interest (ROls). The ROIs which appeared in a hypoattenuation area were recorded. Cortical CBF was tested for specificity and sensitivity as a predictor of hypoattenuation by means of a receiver operating characteristic curve. RESULTS: Mean age was 58 (SD +/- 12.4) years. The median Fisher score and Hunt and Hess scale were 2 and 3, respectively. The area under the receiver operating characteristic curve was 0.912 (CI 0.896 to 0.926). The cut-off value for best accuracy was 6 mL/ 100 g/min, with a likelihood ratio of 37. CONCLUSION: The present study suggests a threshold of 6 mL/100 g/ min as a predictor of a new low attenuation area. However, each clinician should choose the most useful threshold according to pre-test probability and the cost/effectiveness ratio of the applied therapies.

Reperfusion of low attenuation areas complicating subarachnoid hemorrhage.

Hypoattenuation areas shown on brain CT scans after subarachnoid hemorrhage (SAH) are believed to be associated with persistent ischemia. The aim of this study was to evaluate regional cerebral blood flow (rCBF) in hypoattenuation areas and its evolution over time by means of Xenon CT (Xe-CT). We enrolled 16 patients with SAH who developed a hypoattenuation area in the middle cerebral artery territory. Patients were studied at time zero (the first Xe-CT), within 24 to 96 hours, and 96 hours after the initial Xe-CT. We analyzed 19 hypoattenuation areas caused by vascular distortion, vasospasm, or post-surgical embolization in 48 Xe-CT studies. Areas of hypoattenuation were divided in 2 groups according to initial rCBF. In the first group (n = 15), rCBF was initially above 6 mL/100 gr/min but only 2 were still ischemic (rCBF < 18 mL/ 100 gr/min) 96 hours after the first Xe-CT, while 7 (58%) were hyperemic. Conversely, in the second group with severe ischemia (rCBF < 6 mL/100 gr/min; n = 4) mean rCBF increased (p = 0.08) but still remained below the ischemic threshold. In severely ischemic lesions, rCBF reperfusion occurs but is probably marginally relevant. Conversely, in lesions not initially severely ischemic, residual CBF gradually improved and frequently became hyperemic. The functional recovery of these zones remains to be evaluated.

Cerebral blood flow in mean cerebral artery low density areas is not always ischemic in patients with aneurysmal subarachnoid hemorrhage--relationship with neurological outcome.

Aneurysmal subarachnoid hemorrhage (SAH) can be complicated by reduction of regional cerebral blood flow (rCBF) from large conductance vessels leading to focal edema appearing as an area of hypoattenuation on CT. In this study we included 29 patients with SAH due to aneurysmal rupture, having 36 CT low density areas within the middle cerebral artery territory in whom a total of 56 Xenon-CT (Xe-CT) studies were performed. Collectively, we evaluated 70 hypoattenuated areas. rCBF levels were measured in two different regions of interest drawn manually on the CT scan, one in the low density area and the other in a corresponding contralateral area of normal-appearing brain tissue. In the low density area (22.6 +/- 22.7 ml/100 gr/min) rCBF levels were significantly lower than in the contralateral area (32.8 +/- 17.1 7 ml/100 gr/min) (p = 0.0007). In the injured areas deep ischemia (CBF < 6 ml/ 100 g/min) was present in only 25.7% of Xe-CT studies, suggesting that hypodense areas are not always ischemic, whereas in 43.7% of the lesions/Xe-CT studies we found hyperemic values. Patients with a better outcome had hyperemic lesions, suggesting brain tissue recovery in injured areas.

Cerebral blood flow mapping in two different subtypes of intraparenchymal hemorrhagic traumatic lesions.

The pathogenesis and the viability of edematous tissue may be different in traumatic hematomas and traumatic contusions. We tested the hypothesis that mapping of regional Cerebral Blood Flow (rCBF) was different in these two subtypes of traumatic intraparenchymal lesions. We evaluated rCBF by means of Xenon-enhanced computerized tomography (Xe-CT) in 59 traumatic intracerebral lesions from 43 patients with severe head injury. One-hundred-nine intracerebral lesions/Xe-CT CBF measurements were obtained. The rCBF was measured in the hemorrhagic core, in the intralesional oedematous low density area and in a 1 cm rim of apparently normal perilesional parenchyma of both lesion subtypes. Not statistically significant lower rCBF levels were found in the edematous area of traumatic contusions. In traumatic hematomas rCBF levels were lower in the core than in the low density area, suggesting that rCBF in edematous area is marginally involved in the initial traumatic injury and that edema is probably influenced by the persistence of the hemorrhagic core. Conversely, in the traumatic contusions a difference in rCBF values was found between core, low density area and perilesional area, indicating that rCBF of the low density area is related to a concentrical distribution of the initial injury.

Mixed dishomogeneous hemorrhagic brain contusions. Mapping of cerebral blood flow.

The aim of the study was to verify whether regional cerebral blood flow (rCBF) was distributed centrifugally in traumatic hemorrhagic contusions with multiple cores within an oedematous area. Seventeen traumatic brain contusions, from 14 patients with severe head injury (GCS < 9), were analyzed during 39 Xenon-enhanced computerized tomography (Xe-CT) studies. The CBF was measured in 3 concentric regions of interest (ROls): the hemorrhagic core, the intracontusional oedematous low density area and a 1 cm rim of pericontusional normal-appearing brain tissue surrounding the contusion. Differences between rCBFs in the three ROIs were found (p < 0.0001). rCBF in both the hemorrhagic core (21.4 +/- 19.4 ml/ 100gr/min) and the intracontusional low density area (28.4 +/- 19 ml/100gr/min) were lower than rCBF in pericontusional normal-appearing area (41.9 +/- 16 ml/100gr/min) (p < 0.0001). No significant differences were found between rCBF measured in the hemorrhagic core and intracontusional low density area (p = 0.184). Our study suggests that in the mixed density contusions with multiple hemorrhagic cores, the CBF is concentrically distributed, improving from the core to the periphery.

Induced acute arterial hypertension and regional cerebral flow in intracontusional low density area.

Traumatic brain contusions have been associated with regional ischemia. We aimed to measure the effect of induced supra-normal values of cerebral perfusion pressure (CPP) on regional cerebral blood flow (rCBF) in the intracontusional low density area surrounding the contusional hemorrhagic core. In 7 severely head injured patients (GCS < or = 8) harbouring a contusion larger than 2 cm, the rCBF levels were measured, by means of Xenon-enhanced CT, in: 1) the intracontusional low density area: 2) contralaterally, in a normal brain symmetric area. CBF studies were performed at a baseline CPP of 65.3 mmHg +/- 7 and after 20 minutes of norepinephrine-induced CPP supernormal values (88.3 mmHg +/- 10.5) (p = 0.0013). A "paradoxical" reduction of rCBF levels was observed in both the intracontusional low density area (p = 0.07) and the contralateral "normal" area (p = 0.08). In particular, this decrease of rCBF in the intracontusional low density area (-25.7 + 10 ml/100gr/min) (p = 0.0009) was present in only 4 cases, having a mean rCBF at baseline of 25 +/- 16 ml/100gr/min. In the remaining 3 cases in which rCBF at baseline was abnormally low (12 +/- 7 ml/ 100gr/min), rCBF values improved slightly (3.6 +/- 2 ml/100gr/min) (p = 0.61). An acute increase of CPP seems to marginally affect rCBF in the intracontusional low density area having critically reduced initial values, but may greatly reduce rCBF in subjects starting from non-critical baseline values.