Residual Cerebrovascular Morbidity Despite Treatment in Pediatric Sickle Cell Disease Highlights Opportunities for Earlier, Intensified Monitoring and Treatment

Abstract

We read with great interest the article by Casella et al. [1] which reported the results of the HU Prevent Trial. Regardless of the limitations acknowledged by the authors, Casella's study suggests that Hydroxyurea (HU) may have neuroprotective effect in very young children with sickle cell disease (SCD) and support a phase III study to encourage the early use of HU in all infants with SCD. While many advances have been made in the understanding of the pathophysiology and the identification of the cornerstones of comprehensive care to reduce mortality in the first years of life, cerebral vasculopathy remains a cause of acute mortality and long-term morbidity with impact on quality of life [2-4].

Here, we report on cerebral vasculopathy morbidity in a longitudinal pediatric SCD natural history cohort in a high-income country, showing both the benefit of HU in reducing cerebral vasculopathy but also the need to intensify treatments in young children in order to reduce the serious impact of cerebral complications later in life. We performed a retrospective analysis using data from children with SCD who, at age of diagnosis, were prospectively enrolled in the SCD Natural History Study, which began in December 1, 2007 and is an ongoing, longitudinal follow-up study recording information on the physical examination, laboratory parameters, acute and chronic complications during every outpatient and inpatient visit since diagnosis. The censoring date was December 31, 2020, or the date of death, enrollment in a randomized clinical trial assessing new therapeutic drugs, transfer to a new center, or Hematopoietic Stem Cell Transplantation (HSCT). The study was approved by the Research Ethic Committee of Padua University Hospital, and written informed consent was obtained by each caregiver. The Italian Association of Pediatric Hematology Oncology (AIEOP) guidelines were used to guide acute event treatment, monitoring of chronic organ damage and standardize treatment with Hydroxyurea (HU), red blood cell transfusion and HSCT, as well as management of acute and chronic complications (6–7). HU was indicated in case of more than two vaso-occlusive-crisis (VOC) per year or chronic anemia with hemoglobin < 8 g/dL. Regarding cerebral vasculopathy, the standard care included annual Transcranial doppler (TCD) from age 2 and Magnetic Resonance Imaging/Magnetic Resonance Angiography (MRI/MRA) from 6 years, every 2 years, or earlier in case of abnormal TCD. In this analysis, we evaluated event-free survival (EFS) and cumulative incidence analysis of a composite neurological event defined as the occurrence of at least one of the following: clinical neurological event (transient ischemic attack or stroke), abnormal/conditional TCD, new stenosis at MRA, new silent cerebral infarct (SCI) at MRI.

Fisher and chi-square tests were used to compare categorical variables. Kaplan–Meier method was used to estimate survival probabilities, and heterogeneity among strata was tested with log-rank test. EFS was calculated from diagnosis to the time of the first neurological event, of HSCT, of death or loss to follow-up. The SAS software version 9.4 was used to perform all statistical analyses.

A total of 169/182 (92.9%) patients enrolled in the natural history cohort had at least one neuroimaging exam (110 patients with MRI/MRAs for a total of 288 MRI/MRAs; 155 patients with TCDs with a total of 583 TCDs), 91 (50%) had both TCD and MRI/MRA. At the censoring date, the median follow-up was 79.8 months (range: 2.1–298.6 months; interquartile range [IQR]: 36.9–126.3 months), corresponding to 1358.44 patient-years of follow-up. Demographics are shown in Table S1. Median age at diagnosis was 24.4 months (IQR 8.4–56.5 months) due to lack of a national newborn screening program; nevertheless, 47% were diagnosed before 24 months of age. As of 31 December, 2020, three had been lost to follow-up and four deceased (3 due to Pneumococcus septic shock, 1 for a trauma accident). During follow-up, 108 children (64%) received HU starting at a median age of 69.5 months (IQR 39.3–109.3). Age at diagnosis was different between the HU-treated children compared to the non-treated: median 21.9 (IQR 6.6–42.7) versus 38.2 months (IQR 17.9–66.7).

A total of 49/169 (29%) patients had neurological complications, with a cumulative incidence of neurological complications 5 years after diagnosis of 24.5 (95% CI, 18.3–31.2) and EFS of 80.9 (95% CI 73.9–86.3). The incidence of neurological complications was lower in those who received HU within 5 years of diagnosis with cumulative incidence of 11.7 (95% CI 6.3–18.7) versus 45.1 (95% CI 32.5–56.8) (p < 0.0001) (Figure 1) and EFS of 89.1 (95% CI 81.2–93.8) versus 65.6 (95% CI 51.2–76.6) (p < 0.0001). However, 11% of the children on HU still presented significant neurological complications 5 years after diagnosis in spite of undergoing HU, thereby showing a need of more intensive treatment to reduce the cerebrovascular complications.

FIGURE 1

Open in figure viewerPowerPoint

(A) Cumulative incidence of neurological events in children with Sickle Cell Disease (SCD); (B) Cumulative incidence of neurological event according to Hydroxyurea treatment in children with sickle cell disease (SCD).

Abnormal and conditional TCD velocities were found in 1.6% and 10.9%, respectively, only in SS/Sβ° patients. Detailed results of TCD analysis were previously described [5].

For the 110 children that had at least one MRI/MRA, the median follow-up was 113.9 months (range 6.2–301.0) (IQR 79.2–153.3). More than one MRI was performed by 74 children (range 1–16). Median age at time of first MRI was 6.52 years (range 1.04–20.64) (IQR 4.64–10.01). Median time between the first and the last MRI in patients with at least 2 exams was 4.8 years (range 0.3–15.7). SCIs occurred in 37 patients (33.6%), the majority of them (29/37, 78%) presenting SCIs already at the first MRI. The mean lesion burden at first MRI was 1550.1 mmc (range 60.5–5916.0). Cumulative Incidence of SCIs at 5 years from diagnosis was 22.7% (95% CI 15.2–31.1). In our cohort, SCI prevalence was similar among genotypes (p = 0.999): 5/14 (35.7%) in the HbSC/Sβ+, 32/96 (33.3%) in the HbSS/Sβ°. All SCI in the HbSS/HbSbeta° group were observed in patients with HbSS (32/89). No difference in neurological complication was present at baseline between children who were then treated with HU compared to those who were not (Table S2). Cumulative incidence according to treatment received before the SCI is presented in the Figure S1. Significant difference in SCI 5-year Cumulative Incidence was observed between patients receiving treatment compared to those who did not receive treatment, with the lowest incidence in the group that was receiving HU (5 year Cumulative Incidence of 63.4 vs. 10.5 vs. 1.9, p < 0.001 in those without therapy vs. those in combined treatment vs. those in HU, respectively). Two children who developed SCI were on chronic transfusion for abnormal/high conditional TCD (n.1 patient) or Moya-Moya Syndrome with multiple stenosis (n.1 patient); five children had experienced combined treatment (2 for history of abnormal/conditional TCD and stenosis at MRA, 3 for persistence of VOC while on HU); five were on HU.

Median age at the start of HU was 81.5 months (IQR 45.9–133.5). Ten patients were already receiving HU before diagnosis of SCI: median time from the start of HU to detection of SCI was 55.6 months (IQR 20.8–82.9).

A “new event” was defined as the appearance of a SCI that wasn't present at the previous MRI or an increase in lesion burden of a previously detected SCI. A total of 45 new events were detected: 31 patients had one event only, 6 patients had more than one event (2 events: n = 5; 4 events: n = 1). Mean lesion burden for all the new events was 2718.35 (range 30.0–61734.0).

Our data from a real-world cohort in a high-income country, where newborn screening is absent, but comprehensive care for SCD is free and delivered in high specialized centers with ensured chronic organ damage monitoring, shows significant neurological complications, particularly SCIs, in spite of a median age at diagnosis of SCD of 24.4 months and a median age at first MRI of 6.52 years. Most patients already have SCIs at the first MRI, confirming previous data in young children [5]. Like other European cohorts, where routine MRI is done once sedation is unnecessary, cumulative incidence is high and can progress. Given that cumulative risk of SCI can be as high as 39.1% by age 18 [2] and can further increase over time and considering that baseline SCI predicts progressive infarcts and vasculopathy [6], the need to avoid the development of cerebral vasculopathy and especially SCI since early childhood is crucial. In our cohort, children on treatment have a lower risk to develop overall cerebral vasculopathy both as a composite endpoint, as well as SCI, underlining the potential benefits of early preventive treatment. Moreover, our data support trials for early screening and intensification of treatment on a broader scale, suggesting that optimization of therapy, weather of single or combination of drugs, might be beneficial. Sedation issues can be present in infants who must undergo an MRI, but it is probably time to consider innovative approaches in order to perform neuroradiological exams in young children with SCD.

Long-term data collection in our high-income setting indicates that, although being offered highly specialized monitoring and treatment options, a subgroup of young children still displays significant cerebrovascular complications and would require intensified treatment at an early age. It is important to overcome the neglect that young children have experienced in access to therapeutic options by prioritizing trials with drugs for the prevention of cerebrovascular complications in children below 5 years of age, especially considering the permanent damage that similar complications present.