TWiTCH

Background

Phase III open-label non-inferiority RCT. 121 children aged 31 months to 18 years (median 8 years) with SCD (HbSS or HbSβ0) who had received chronic transfusion for ≥18 months due to prior TCD ≥200 cm/s and had achieved TCD normalization (<170 cm/s on two consecutive measurements). No history of overt stroke. Conducted at 26 STOP and additional North American sites. Designed to test whether hydroxyurea could replace chronic transfusion for primary stroke prevention while reducing iron overload.

Interventions and follow up

Arm A: Transition to hydroxyurea (started 10 mg/kg/day, escalated every 8 weeks to MTD ~26 mg/kg/day; concurrent gradual transfusion taper over 12 months)
Arm B: Continue chronic transfusion (target HbS <30%)
Primary endpoint: TCD velocity at 24 months (non-inferiority margin: 15 cm/s, upper bound of one-sided 97.5% CI)
mFollow up: 24 month

Results

TCD velocity: Mean 143 cm/s (HU) vs 138 cm/s (transfusion); difference 4.54 cm/s (95% CI −1.84 to 10.92) — non-inferior (upper CI 10.92 < 15 cm/s margin)
Ferritin change: −950 ng/mL (HU) vs stable (transfusion) — significant iron reduction
Strokes: 0 in either arm during study period

Adverse events

Main adverse events: HU arm: grade ≥3 neutropenia in 5%, mild reticulocytopenia. One HU patient discontinued due to rash. Transfusion arm: continued alloimmunization risk, ongoing iron loading. No treatment-related deaths. Iron overload substantially improved in the HU arm with ferritin reduction of ~40% over 24 months.

Conclusions

Hydroxyurea was non-inferior to chronic transfusion for maintaining normal TCD velocity in high-risk SCD children who had previously normalized on transfusion, while substantially reducing iron overload, establishing hydroxyurea as the preferred transition strategy once TCD normalizes.

Key Limitations

Key Limitations: Open-label design. TCD velocity is a surrogate for stroke, not actual stroke occurrence. Eligible only for children who had already achieved TCD normalization on transfusion — those who fail to normalize are excluded, limiting generalizability. Prior stroke patients were excluded. Long-term (>24 months) stroke outcomes after transition are unknown. Hydroxyurea dose escalation protocol required frequent visits not available in all settings.

Clinical Context

TWiTCH established that gradual transition from transfusion to hydroxyurea is safe and effective once TCD normalizes. This is now the standard approach, recommended by ASH 2020 SCD guidelines. Real-world data confirm stroke rates remain low on HU post-transition. Children with prior stroke or those who fail TCD normalization on transfusion continue to need chronic transfusion. Exa-cel (Casgevy) and other curative approaches may eventually supplant long-term HU in eligible patients.

References

References: Ware RE et al, Lancet 2016 (TWiTCH primary)