PREDICTED LIFE EXPECTANCY IMPROVES CLINICAL INTERPRETATION OF SURVIVAL MODELS IN SICKLE CELL DISEASE

@article{moura2026predicted,
  title = {{PREDICTED LIFE EXPECTANCY IMPROVES CLINICAL INTERPRETATION OF SURVIVAL MODELS IN SICKLE CELL DISEASE}},
  author = {Isabel Cristina Gomes MOURA and Carla Luana DINARDO and Mina Cintho OZAHATA and Sheila MATEOS and Márcio OIKAWA and Alessandra FERRAZ and Luiz AMORIM and André BELISáRIO and Anna Bárbara PROIETTI and Shannon KELLY and Brian CUSTER and Ester SABINO and Vanderson ROCHA},
  journal = {Hematology, Transfusion and Cell Therapy},
  year = {2026},
  month = oct,
  doi = {10.1016/j.htct.2026.106694},
  url = {https://doi.org/10.1016/j.htct.2026.106694},
  note = {CC BY},
}

TY  - JOUR
TI  - PREDICTED LIFE EXPECTANCY IMPROVES CLINICAL INTERPRETATION OF SURVIVAL MODELS IN SICKLE CELL DISEASE
AU  - Isabel Cristina Gomes MOURA
AU  - Carla Luana DINARDO
AU  - Mina Cintho OZAHATA
AU  - Sheila MATEOS
AU  - Márcio OIKAWA
AU  - Alessandra FERRAZ
AU  - Luiz AMORIM
AU  - André BELISáRIO
AU  - Anna Bárbara PROIETTI
AU  - Shannon KELLY
AU  - Brian CUSTER
AU  - Ester SABINO
AU  - Vanderson ROCHA
PY  - 2026
DA  - 2026/10/23
JO  - Hematology, Transfusion and Cell Therapy
DO  - 10.1016/j.htct.2026.106694
UR  - https://doi.org/10.1016/j.htct.2026.106694
AB  - Introduction: Risk prediction in sickle cell disease (SCD) is commonly reported using hazard ratios, which are often difficult to interpret in clinical practice. Estimating the adjusted predicted life expectancy from parametric survival models may provide a more intuitive measure of prognosis and facilitate communication of disease burden to clinicians and patients. Objective: To estimate genotype-specific predicted life expectancy associated with major clinical risk factors in patients with SCD using standardized Weibull survival models. Methods: This longitudinal study included 2,696 participants from the Brazilian Recipient Epidemiology and Donor Evaluation Study III (REDS-III), a prospective multicenter cohort of patients with SCD, comprising 2,052 individuals with HbSS/HbSβ⁰ and 644 with HbSC disease. Parametric Weibull survival models accounting for left truncation were fitted separately for each genotype. For each prognostic factor, adjusted predicted life expectancy was estimated using marginal standardization by assigning all individuals to each exposure level while preserving the observed distribution of adjustment covariates. Confidence intervals were obtained by patient-level bootstrap resampling. Associations were assessed using likelihood ratio tests. Results: Among the 2,696 participants, sex distribution and age at enrollment were similar between HbSC and HbSS/HbSβ⁰ groups (54.2% vs. 52.7% female, p=0.516; mean age 20.7±15.8 vs. 20.0±13.2 years, p=0.330). However, HbSC participants were more frequently ≥40 years at enrollment (14.9% vs. 9.6%, p<0.001). Distinct genotype-specific prognostic profiles were identified. Among individuals with HbSS/HbSβ⁰, pulmonary hypertension was associated with the greatest reduction in adjusted predicted life expectancy, corresponding to nearly 10 fewer years of life (26.4 vs. 36.3 years; difference -9.9-years, 95% CI -15.5 to -4.1; p=0.003). Likewise, patients requiring >0.33 transfusions/year had a markedly shorter predicted life expectancy than those with a lower transfusion burden (30.2 vs. 41.1 years; difference -10.8-years, 95% CI -18.4 to -5.1; p<0.001). Acute chest syndrome and hospitalization burden (>1 hospitalization/year) showed non-significant trends toward lower reduced life expectancy, whereas hydroxyurea use was not associated with predicted age at death. In contrast, among HbSC patients, hydroxyurea use identified a subgroup with markedly lower adjusted predicted life expectancy (34.6 vs. 50.8 years; difference -16.3 years, 95% CI -38.1 to -1.7; p=0.029), while annual hospitalization burden and acute chest syndrome were not significantly associated with survival. Discussion and Conclusions: Expressing prognosis as predicted life expectancy rather than hazard ratios provides an intuitive and clinically meaningful measure of prognosis in SCD. Prognostic determinants differed substantially by genotype. In HbSS/HbSβ⁰ disease, with pulmonary hypertension and higher transfusion burden were associated with approximately 10 fewer years of predicted life, highlighting patients who may benefit from closer monitoring and earlier disease-modifying interventions. In HbSC disease, hydroxyurea use identified a higher-risk subgroup, likely reflecting treatment indication in patients with more severe disease. This approach translates complex survival models into clinically interpretable estimates that may improve patient counseling, individualized risk stratification, and clinical decision-making.
SN  - 2531-1379
SN  - 2531-1387
ER  -