Screening Brain MRI in Sickle Cell Patients: An Insufficiently Implemented Intervention with Critical Importance

Background:

The hallmark of sickle cell disease (SCD) is vaso-occlusion and multi-organ ischemic damage. Neurological dysfunction can manifest as difficulties in attention, memory, executive functioning, and processing speed affecting the patient’s scholarly and academic progress. Screening and early intervention are critical. Although ASH guideline panel 2020 suggests at least one time screening brain MRI in SCD adults, small observational studies suggest that this is not well implemented in clinical practice.

Methods:

A retrospective chart review of 93 patients following at the UT Comprehensive Sickle Cell Center in Houston, Texas between January and April 2024. 435 patients were screened, but only patients with brain MRI were selected for the analysis. We could not find other records of brain MRI using extensive search methods to review EPIC EMR and other hospital records, and/or there was no documentation of brain imaging. Some patients could have had imaging in another health system, and we were unable to access the report.

Demographic information and MRI results were reported. Laboratory values including mean hemoglobin, hemolysis parameters, creatinine and ALT were collected to assess disease severity. Data from the overall cohort was analyzed using standard descriptive statistics. This study aims to identify the various findings of brain MRI in those patients and the differences between those patients and others with normal imaging.

Demographic Data and Lab Results:

This study comprised of 93 SCD patients, the average age being 37.47 ± 14.59 years. 65% were females and 35% were males. 73% (n=68) had HbSS genotype, 22% (n=20) had HbSC genotype, 4% (n=4) had HbS/beta null genotype and 1% (n=1) had HbS/beta plus genotype. 99% of patients were African American and 1% were Hispanic. 65% were receiving disease-modifying therapy (DMT). Of those, 53% (n=49) used hydroxyurea, 14% (n=13) used voxelotor, 5% (n=5) used crizanlizumab, and 2% (n=2) were on L-glutamine. 34% (n=32) of patients were currently on or had previously received chronic exchange transfusion.

In patients with an abnormal MRI, the average mean hemoglobin (g/dl) was 8.97 ± 1.63 and total bilirubin (mg/dl) was 2.74 ± 2.59. The average reticulocyte count (%) was 9.01 ± 6.13 and average LDH (U/L) was 451.39 ± 228.14. Mean creatinine (mg/dl) was 0.77 ± 0.41 and ALT (U/L) had an average of 29.88 ± 29.57. 66.7% (n=46) were receiving one or more DMT.

In patients with normal MRI, the average hemoglobin (g/dl) was 9.45 ± 1.60 and total bilirubin (mg/dl) averaged 2.12 ± 1.57. The mean reticulocyte count (%) was 7.19 ± 5.19 and average LDH (U/L) was 324.87 ± 155.75. The average creatinine (mg/dl) was 0.77 ± 0.26. ALT (U/L) had an average of 21.13 ± 14.32. Only 58.3% (n=14) were on a DMT of some kind.

Of the cohort of 93 patients, 69% had a TRV max reported. The average TRV max was 2.38 ± 0.40 m/s. 26% of patients had chamber dilation noted on their echocardiograms. 38% of patients in the overall cohort had a known history of TIA or stroke.

Diagnostic Results:

Out of 435 adult patients with SCD, 21.3% (n=93) had a brain MRI. Of these, 74.1% (n=69) had an abnormal MRI. In this cohort of patients, 24.6% (n=17) had evidence of remote ischemic infarcts. 43.4% (n=30) had reported chronic ischemic changes and encephalomalacia. 8.6% (n=6) had white matter changes. 17.3% (n=14) had bony changes, mostly bone thickening. 4.3% (n=3) had moyamoya disease and 4.3% (n=3) had an underlying aneurysm. 2 patients had a brain hemorrhage, and 4 patients were found to have pituitary changes. Of the patients with an abnormal brain MRI, only 47.7.8% (n=33) had a known history of a TIA/stroke.

Discussion:

To our knowledge, this is the largest retrospective review of adult SCD patients reporting abnormal MRI findings. Only about a fifth of the patients had an MRI brain and most of the imaging was performed for diagnostic purposes rather than screening. In patients with abnormal MRI, less than half had a history of TIA/stroke. Analysis of the laboratory data indicates that these patients have more severe disease suggested by lower mean hemoglobin and higher values of hemolysis parameters. Therefore, screening MRI is of paramount importance in all SCD patients, especially in the sicker population.

This study is ongoing; with a plan to screen 1000 patients in our center, we intend to implement a strategy of one-time screening MRI for all adult SCD patients with subsequent formal referral for neurocognitive testing.