Enhanced MRI Evaluation of Children with Sickle Cell Following Hematopoietic Stem Cell Transplantation

Introduction: Sickle cell disease (SCD) CNS vasculopathy (SCNSV) is a frequent indication for hematopoietic stem cell transplantation (HSCT).  Untreated, SCNSV can be progressive and impair quality of life (QoL) and cognitive functioning.  By clinical MRI/MRA assessment, HSCT is thought to halt progression of SCNSV.  Quantitative analysis of T2-weighted FLAIR MRI for white matter hyperintensity (WMH) can provide a meaningful estimate of small vessel cerebrovascular burden. Adding WMH assessment, we asked whether HSCT for SCD halted long-term progression of SCNSV, including small vessel involvement, despite transplant-associated CNS risks.  QoL assessment can track school functioning and physical, emotional and social functioning.

Methods:  This retrospective single site study compared MRI analyses pre-transplant to 1-7-years post-HSCT.  Subject eligibility required availability of clinical MRI scans from within 2 months pre-HSCT and at 1-year intervals for at least one year post-HSCT.  Interim scans performed for acute clinical indications were not included in the analysis.  MRI scans were evaluated independent to the initial clinical read by one neuroradiologist, and via in-house developed software to quantitate WMH burden.  QoL was completed by parent- and child-report pre-HSCT and annually thereafter using the Peds QL 4.0^TMwith scores 0-100; higher numbers reflected higher QoL.

Results: 25 patients who received HSCT for SCD between 2003-2014 were evaluated.  Median age at HSCT was 9.9 years (1.4-21.9); male:female 18:7; HbSS (17), HbSC (3), HbS-Bthalassemia (5). Donors were related (14) or unrelated (11).  Stem cell sources were: related bone marrow (BM) (10), unrelated BM (5), related cord blood (CB) (3), unrelated CB (6), or related peripheral blood stem cells (1). Eight patients had CNS pathology as the indication for HSCT.  Transplant complications were: PRES (2), stroke (1), graft failure (2), death (1).  Duration of follow-up was: 1 yr (9), 2 yrs (10), >2 yrs (6).  Only a minority of patients had normal pre-HSCT MRI (5) and normal MRA (11; 6 had 1-2 stenoses <2mm and were rated as ambiguous).  At 1 to 7 yrs post-HSCT, 5 originally normal MRIs were unchanged, 15 MRIs were stable and 4 were improved at 1 year post-HSCT, without subsequent changes by clinical assessment.  One MRI worsened due to peri-HSCT hemorrhagic stroke.  MRAs were unchanged following HSCT.  Preliminary analysis of 18 patients at pre-HSCT revealed that 5 did not have elevated WMH, while elevated WMH was detected in 13 patients.  WMH remained stable in 16 of 18 patients over subsequent annual assessments.  In the 2 patients with markedly elevated WMH pre-HSCT, values decreased at follow-up, corresponding to the resolution of acute or recent infarction.  Overall pre-HSCT QoL by parent-report (N=19) was 65.3 (SD16.3); while child self-report (N=14) was 67.6, (SD13.1).  After mean follow-up of 3 years, parent-report QoL improved to 78.8 (SD15.8), and self-report to 78.7 (SD15.2).

Conclusions: Most children in this retrospective cohort had MRI abnormalities pre-HSCT, and all but 1 were stable or improved post-HSCT, despite PRES and other potential CNS complications.  Pre-HSCT WMH appeared to be unchanged for most of these patients, while stroke-induced WMH appears to decrease over time, suggesting stable small vessel SCNSV following transplantation.  Overall, by MRI/MRA and by preliminary WMH, HSCT appears to have stabilized large and small vessel SCNSV in all but 1 of 25 children.  While only a modest number of patients were assessed, WMH was reproducible at multiple annual time points.  Long-term parent and self-reported QoL indicated improvement from HSCT. Chronically transfused SCD patients could not be compared due to lack of annualized assessment. Future transplant protocols will include enhanced MRI-based tracking, in conjunction with QoL and neuropsychological assessments. These data could be useful for decision-making about SCD transplantation.