Reversal of Cerebral Arteriopathy Post-Hematopoietic Stem Cell Transplant for Sickle Cell Disease

A now 11-year-old male had been diagnosed with severe hemoglobin SS disease by the state newborn screening test. The patient required numerous hospitalizations for SCD-related complications, including multiple episodes of acute chest syndrome, vaso-occlusive pain episodes, and fevers. His baseline hemoglobin concentration while well was 8 g/dL. Initial transcranial Doppler (TCD) screening study performed at 2 years of age was normal, but no subsequent TCD studies were obtained despite recommendations by providers. At 5 years of age, the patient was started on hydroxyurea but his adherence to treatment was inconsistent. At the age of 6 years, he presented with fever, left leg pain, and acute left-sided upper and lower extremity weakness (National Institutes of Health stroke scale score 2). His hemoglobin concentration was 7 g/dL. Emergent management was initiated with intravenous crystalloids, oxygen, morphine, and simple transfusion. MRI of the brain revealed multiple acute subcortical infarctions in the right middle cerebral artery territory. Magnetic resonance angiography (MRA) showed findings consistent with a moyamoya pattern of cerebral large vessel arteriopathy (Fig 1A). A partial exchange transfusion was performed the next day. Once the patient was afebrile, a high flow double lumen central venous access device was placed, and a chronic monthly exchange transfusion program was started to maintain a circulating hemoglobin S level <30%. Single agent antiplatelet therapy with low-dose aspirin was also started. One month after the patient’s initial stroke presentation, catheter-directed angiography documented severe stenoses of the right carotid terminus, and proximal right M1 and proximal right A1 with hypertrophic medial and lateral lenticulostriate collateral vessels (moyamoya) (Fig 2A, C). Diffuse stenosis of the left A1 and transdural vascularization of the left frontal lobe by the left middle meningeal artery were also present. Five months after his initial stroke presentation, the patient underwent uneventful indirect right superficial temporal artery to right middle cerebral artery bypass with right frontal and right parietal burr holes. Five months after revascularization surgery, the patient had a brain MRI and MRA which indicated arterial stenoses were stable and the bypass graft was patent with improved perfusion of the right cerebral hemisphere.

Given the patient’s ischemic stroke, he met criteria for HSCT as definitive curative therapy. Roughly 18 months after the stroke, the patient underwent an unrelated donor-matched (8 out of 8 allele-matched human leukocyte antigens) HSCT. Other than his SCD-related cerebrovascular disease complications, his only other pretransplant comorbidity was transfusion-related iron overload. Pre-HSCT brain MRI and MRA imaging showed expected evolution of subcortical ischemic changes and no progression of intracranial large vessel arteriopathy. The patient had a reduced intensity conditioning regimen with alemtuzumab, fludarabine, melphalan, and thiotepa. He achieved neutrophil engraftment on day 13 and platelet engraftment on day 18 post-HSCT. His post-HSCT course was complicated by fevers of unknown origin, which resolved with glucocorticoids after extensive infectious disease evaluation. His graft-versus-host disease prophylaxis regimen was calcineurin inhibitor-based and he remained on full dose immunosuppression until 1 year post-HSCT. He has had no evidence of graft-versus-host disease and 100% donor whole blood, and T cell-sorted chimerisms to date. A repeat brain MRI/MRA 1 year after HSCT showed stable large vessel cerebral arteriopathy.

Three years after HSCT, catheter-directed angiography, performed to assess disease progression and bypass maturation, showed a striking reversal of steno-occlusive changes involving the right carotid terminus, right M1, right A1, and left A1 (Fig 2B, D). This study also documented significant improvement in perfusion of the right cerebral hemisphere and left anterior cerebral artery distribution (Fig 3). Although concurrent MRA obtained just before catheter-directed angiography also suggested improvement in the patient’s cerebral large vessel arteriopathy (Fig 1B), the findings are more subtle.

Overall, the patient has been well since discontinuing immunosuppression. He remains on daily single-agent antiplatelet therapy with low dose aspirin. He continues to have a left hemiparesis, which, although much improved, causes mild persistent left foot drop requiring a daytime articulated ankle foot orthotic with dorsiflexion assist. Chronic Achille’s tendon contractures are managed with a night brace to induce stretching. At 5 years follow-up, the pediatric stroke outcome measure is 1.¹² 

This report demonstrates a dramatic reversal of symptomatic SCD-associated cerebral large vessel arteriopathy by gold-standard, catheter-directed cerebral angiography in a child treated by allogenic HSCT. In our patient, severe steno-occlusive arterial lesions and associated moyamoya collateral vessels almost completely resolved over a period of 3 years. Although previous reports have indicated stabilization and even improvement of SCD-associated cerebral large vessel arteriopathy post-HSCT on the basis of clinical findings, transcranial Doppler velocities, noncontrast MRA and MRI, and biomarkers of cerebral perfusion and oxygenation, such changes may be attributable to alterations of hematologic parameters alone rather than an alteration of arterial structure and morphology.^10,11,13,14  Our report demonstrates a normalization of cerebral arterial structure and morphology as assessed by catheter-directed angiography after HSCT.

There is some previous evidence of neuroimaging demonstrating post-HSCT stabilization of neurovascular changes associated with SCD. A French cohort of 87 pediatric SCD patients who underwent matched related or mismatched related myeloablative HSCT between the years of 1988 and 2004 revealed there were no new ischemic lesions detected after engraftment and cerebral arterial velocities were significantly reduced post-HSCT at a mean follow-up time of 6 years.¹⁰  Another retrospective study from Saudi Arabia showed that 15 pediatric patients with SCD had stable cerebral arteriopathy findings on post-HSCT MRI and MRA at a mean follow-up time of 52 months.¹¹  Unlike gold-standard, catheter-directed angiography, MRA is prone to artifacts and sensitive to differences in technique and scanners. Notably, the MRA findings in our patient were reported as stable, even though contemporary, catheter-directed cerebral angiography showed a striking improvement. Arguably, improvement was evident on the MRA but was not reported, perhaps given the lack of confidence in MRA as a definitive diagnostic tool (Fig 1). This phenomenon may had led to underreporting of improvements in cerebral arteriopathy in previous studies where cerebrovascular imaging follow-up is limited to MRA. Although others have described reversal of cerebral arterial stenoses on MRA after HSCT in pediatric SCD patients, confirmation on gold-standard, catheter-directed angiography has not been previously reported.¹⁵ 

A recent prospective study of 10 pediatric patients with SCD who underwent HSCT showed normalization of their cerebral hemodynamics after HSCT.¹³  Cerebral hemodynamics were assessed using cerebral blood flow and fractional oxygen extraction as measured by MRI 1 to 3 months before HSCT and 12 to 24 months after HSCT.¹³  The 10 patients in the cohort had post-HSCT cerebral blood flows and fractional oxygen extractions comparable to non-SCD control patients and improved cerebral hemodynamics compared with children with SCD receiving chronic red blood cell transfusion therapy after a scheduled transfusion.¹³  After controlling for hemoglobin, it was shown that HSCT and transfusion achieved similar reductions in cerebral blood flow and fractional oxygen extraction. The authors proposed that the main advantage of HSCT over chronic transfusion therapy is the long-term normalization of hemoglobin.¹⁶  Additionally, using MRI and diffusion-tensor imaging, Brazilian investigators reported improvement in white matter lesions associated with SCD post-HSCT in 28 patients.¹⁷  These results, and the findings reported here, provide strong cumulative evidence that curative therapy with HSCT has the potential to normalize cerebral perfusion, cerebral oxygenation, and cerebral arterial structure in children with symptomatic, SCD-related cerebral large vessel arteriopathy.

It is important to note that, although our patient had cerebral revascularization surgery before HSCT, such surgery has been extensively studied and is not known to reverse the underlying steno-occlusive lesions in patients with cerebral arteriopathies. These surgeries improve cerebral perfusion by providing alternate routes of cerebral blood flow that bypass steno-occlusive lesions.¹⁸  Additionally, it is also worth noting that the normalization of cerebral arterial structure and morphology observed in our patient began after 6 years of age, when normal cerebrovascular growth and maturation would have been nearly complete.¹⁹  This implies that a postmaturational remodeling process occurred after HSCT. Hence, some reversal of cerebral arteriopathy may occur in older children and adults.

The resolution of large artery stenoses after HSCT as documented by angiography in our patient is most likely the result of chronic vascular remodeling promoted by the elimination of in situ proinflammatory endothelial injury processes including rheometric shear stress and oxidative stress. A contribution of vasomotor tone effects, particularly in the small resistance vessels of the brain, seems likely given that changes in oxygenation have the greatest impact on the vasomotor tone of small resistance vessels. Partial and even complete reversal of arterial stenoses because of inflammatory arteriopathy have been reported after HCST and immunotherapy in other settings.^20,21  Marked improvement in the luminal diameter of the great vessels has been shown as early as 60 days after HCST in Takayasu’s disease.²⁰  In a patient with moyamoya arteriopathy caused by the proinflammatory effects of homozygous sterile alpha-motif and histidine–aspartate domain-containing protein 1 mutation, progressive reversal of cerebral arteriopathy occurred over a 4-year period after initiation of tocilizumab therapy.²¹  Although the results of HCST in our patient show that some SCD patients with intracranial large vessel arteriopathy can experience reversal of large vessel stenoses, it is unlikely that a chronically occluded artery will be restored in a patient with advanced forms of disease (Suzuki stage 4–6). Further study is warranted to understand more fully the determinants of intractability in cerebral arteriopathy patients undergoing HCST.

In conclusion, the remarkable reversal of angiographic arteriopathic changes and associated normalization of cerebral hemodynamics after successful HSCT in our SCD patient suggest the possibility that the structural changes of cerebral large vessel arteriopathy may be reversible. Further study is needed to confirm our findings and to understand the role of HSCT in the treatment of children with SCD-related cerebrovascular disease.

HSCT

hematopoietic stem cell transplant

MRA

magnetic resonance angiography

SCD

sickle cell disease