The pathology of sickle cell disease results from vasoocclusion caused by blockage of the microcirculation by non-deformable red cells filled with polymers of deoxy-hemoglobin S. The disease is characterized by chronic vasocclusive injury to several organs and acute events that include painful crises (which largely involves marrow infarct), acute chest syndrome, acute splenic sequestration, acute sickle liver, cerebral infarcts, renal papillary necrosis, femoral and humeral head necrosis, acute retinal vasoocclusion, and other manifestations. Finding objective criteria to evaluate risk factors and treatment protocols for sickle cell disease is a long-standing goal. To date, MRI of painful crisis has yielded results which are intriguing but difficult to evaluate. Detection of hypoxia may yield a useable marker. These experiments were designed to detect regions of hypoxia which may occur under ambient conditions in patients with sickle cell disease. The image intensity in T2 and T2* weighted images is influenced by the presence of deoxygenated hemoglobin because deoxygenated hemoglobin has a different magnetic susceptibility than oxygenated hemoglobin or tissue. The difference in magnetic susceptibility between the red cells and plasma and between the intra- and extravascular spaces results in more rapid loss of water proton transverse magnetization and ultimately results in reduced image intensity in spin-echo or gradient echo pulse sequences. The change in oxygen saturation in sickle cell patients in brain and bone marrow was investigated while the patient was breathing room air or high oxygen through a face mask and comparing the signal intensity changes in T2 or T2*-weighted images. These results were contrasted to results for normal human controls. A goal of these studies is to separate acute, ongoing events from the results of previous episodes of ischemia and fibrosis or expansion of hematopoietic tissue. Using magnetic resonance imaging (MRI) which is blood oxygenation level dependent (BOLD), we found that transgenic mice with human `- and S-globin have higher levels of deoxyHb in brain, liver, and kidney (areas with pathology) than control mice. We used this technique to examine the level of deoxyHb in the brain and bone marrow of sickle cell patients. In the brain, nine SCD patients and seven normal individuals were scanned while they were breathing first room air and then 100% O2 through a face mask. All SCD patients showed significant global increases in signal intensity caused by O2 inhalation (+5.7%). In contrast, all normal controls showed small or negligible positive change in intensity (+0.57%), which was in agreement with what found in normal subjects reported previously. The intensity change averaged over the brain area scanned for each patient and normal control in response to O2 was measured. While the mean standard deviation of the intensity change obtained on normal subjects is small, there is a large degree of variation in the magnitude among patients. With the only one patient who has had a previously known stroke, the z statistic maps obtained using the pixelwise intensity difference indicates prominently elevated positive z score around the areas of infarct identified in anatomical images. The highlighted region of greater response to O2 also extends beyond the infarcted area medial-posteriolly. O2 saturation recorded with a pulse oximeter throughout the experiments indicated that all sickle cell patients have had resting O2 saturation of less than 94%. We were able to verify in all cases that the O2 saturation increased to 97% or more when the patient was breathing 100% O2 through a face mask. Without any abnormality found in T1 weighted anatomical images, the detected brain signal changes related to inspired oxygen concentration showed significantly higher values than those of normal volunteers. In this study of sickle cell patients and control subjects, we have demonstrated that brain signal changes designed to detect deoxyhemoglobin using MRI can be used to separate patients with SCD and healthy individuals. Furthermore, the experiment showed correlation of the intensity changes in and highlighted the area of infarct in a patient with a history of stroke. Very different patterns emerged in the patients and controls. This technique shows promise for the detection of deoxyhemoglobin in SCA and may help to identify infarcted areas in patients at an early phase. Further studies are needed in order to determine the potential usefulness of this method in evaluating sickle cell patients for the risk of stroke, along with the technical improvements on imaging signal stability over the long period of time required for the studies. we found a mean global change in signal intensity with an increase of 3.7% in SS patients in contrast to a 0.98% increase in normal controls. Signal increases in SS patients were localized to the gray matter and were of a different extent and magnitude for each patient. A high percent deoxyHb under ambient conditions may not imply that the brain is hypoxic because SS RBCs have low oxygen affinity. In bone marrow, we examined sixteen patients both with and without pain femur and hips. Patients with present or past history of hip pain or in steady state were first imaged breathing room air through a face mask. O2 saturation (sat) was monitored by pulse oximetry. O2 then was delivered for two five minute periods and new images were collected and compared, begining at 2 cm above the femoral head and covered a total span of 24 cm in 24 axial slices. Twenty imaging sessions were conducted on 16 SS patients; 8 males and 8 females. Seven sets of images could not be interpreted due to excessive motion. Initial O2 sats were 93% q 3%; final O2 sats were 98-99%. Positive signal, that is presence of deoxyHb under room air conditions, required more than 5 minutes to develop after onset of O2 delivery. DeoxyHb was found predominantly in bone marrow (BM), but was also detected in muscle in patients suffering pain. Nine patients had deoxyHb in BM or muscle, but 4 of these patients had no pain at the time of the examination. Strong positive deoxyHb signal in the femoral heads was found both in patients without cMRI findings of AVN and those with cMRI findings of AVN. Four SS patients without MRI findings of AVN and no pain had no deoxyHb in BM or other tissues. In summary, deoxyHb is detectable in femoral heads during hip pain episodes and occasionally during steady state in patients with a past history of hip pain. Localized areas of deoxyHb are found in some but not all SS patients. In addition, we find that hip pain can be associated with deoxyHb in BM and contiguous muscle mass. We conclude that BOLD-MRI detects areas with high levels of deoxyHb, which can be either ischemic or presumably pre-ischemic. Unlike conventional MRI, it reflects only present pathology. BOLD-MRI is minimally invasive and highly sensitive to areas with deoxyHb, but cannot inform on vascular areas that are not perfused. BOLD-MRI will be useful in further understanding the pathophysiology of VOEs and the evaluation of treatment protocols.
Showing the most recent 10 out of 150 publications
