Molecular characterization of patients with polycythemia vera that develop resistance to hydroxyurea

Hydroxyurea is a cytoreductive agent offered to patients with polycythemia vera (PV) for treating symptoms such as splenomegaly and leukocytosis.¹ Unfortunately, hydroxyurea resistance occurs in 20–30% of patients with PV and is associated with increased risk of thrombosis, disease progression, and adverse survival outcomes. While genomic classification of myeloproliferative neoplasms (MPN) based on data from next-generation sequencing (NGS) has been used for prognosis predictions, the association between molecular characteristics and the risk of developing hydroxyurea resistance is yet to be explored.²

In a letter to the editor, Alberto Alvarez-Larrán and colleagues presented a study in Leukemia that aimed to characterize hydroxyurea-resistant patients with PV by molecular classification and to correlate results with disease progression and survival outcomes.²

Method

• All PV diagnoses were made according to World Health Organization (WHO) criteria, and resistance to hydroxyurea was assessed according to the European LeukemiaNet (ELN) modified criteria.
• Samples were collected and were analyzed by NGS; this included 61 hydroxyurea-resistant patients at time of diagnosis (n = 38) and/or at time of resistance (n = 45).
• A total of 22 cases with paired samples were analyzed at both timepoints.
• Patients who were treated with hydroxyurea and did not develop resistance (n = 59) were used as controls.
• A previously published algorithm was used to hierarchically allocate patients into eight molecular groups.

Results

Baseline patient characteristics are presented in Table 1

Table 1. Patient characteristics and molecular classification at diagnosis in patients with PV according to development of hydroxyurea-resistance versus controls²

• The median follow-up was 7 years from the first hydroxyurea treatment until end of treatment, or death (n = 38). Median follow-up was 6.5 years for hydroxyurea-resistant patients and 7.3 for the controls, p = 0.6.
• The median duration of hydroxyurea therapy was 4.6 years in patients who developed hydroxyurea resistance and 5.9 years in controls.
• The median survival was 15.7 years from the first hydroxyurea treatment (95% CI, 7.6–23.8).
• Correlation of genomic classification with hydroxyurea resistance:
  • Mutations, other than JAK2, were increased in patients with hydroxyurea resistance compared to the controls (Table 1).
  • Patients with hydroxyurea resistance were significantly more likely to have TP53 disruption/aneuploidy (16.4%), and spliceosome/chromatin gene mutations (37.7%), compared to the controls, p < 0.0001 (Table 1).
  • Six of 22 patients with paired samples showed acquisition of molecular markers for hydroxyurea resistance (two patients each for TP53 mutations, ZRSR2 gene mutations, and homozygous JAK2 mutations).
  • After 5 years of treatment, patients with TP53 disruption/aneuploidy had a 64% chance of developing hydroxyurea resistance. This probability was 49% for patients with spliceosome or chromatin aberrations, 27% in patients with homozygous JAK2 mutation, and 14.5% in patients with heterozygous JAK2 mutation, (p < 0.0001 for comparison between all four groups).
• Correlation of genomic classification with outcome:
  • Chances of survival:
    10 years after hydroxyurea treatment, 94% of patients with heterozygous JAK2 mutation, 75% of patients with homozygous JAK2 mutation, 48% of patients with aberrations in spliceosome/chromatin genes, and 59% of patients with TP53 disruption/aneuploidy were still alive (p < 0.0001 for comparison between groups).
  • Risk of thrombosis:
    No statistically significant differences in thrombosis incidence were observed when patients were genetically stratified, except between JAK2 homozygous and heterozygous groups (p = 0.047).
  • Risk of progression to myelofibrosis:
    The rate of progression to myelofibrosis was faster in patients with spliceosome or chromatin gene abnormalities, while it was low in patients with JAK2 (p < 0.0001).
  • Risk of progression to AML/MDS:
    Patients with TP53 or spliceosome/chromatin abnormalities were significantly more likely to progress: nine progressed to acute myeloid leukemia (AML) and three to myelodysplastic syndromes (MDS) (p < 0.0001).

In multivariate analysis, mutations listed in the genomic classification were associated with a higher risk of hydroxyurea resistance (HR, 2.2; 95% CI, 1.5–3.2; p < 0.0001). Also, after correcting for age, patients with TP53 disruption/aneuploidy or spliceosome/chromatin mutations had a higher risk of death when compared with patients with JAK2 heterozygous (HR, 4.2; 95% CI, 1.2–15.1; p = 0.026) and patients with JAK2 homozygous (HR, 2.1; 95% CI, 1.01–4.04; p = 0.046).

Conclusion

In summary, the data support the implementation of genetic classification by NGS for prognosis and clinical decision making in patients with PV. Molecular classification revealed patterns for the risk of disease progression and hydroxyurea therapy resistance for each molecular group, especially for TP53 and spliceosome/chromatin mutations.

This study highlights the need for novel drugs to treat high-risk molecular aberrations, since treatment with hydroxyurea is subject to resistance. Also, this study shows the need for thrombosis prevention in patients with a homozygous JAK2 mutation. In contrast, patients with a heterozygous JAK2 mutation may benefit from long-term treatment with hydroxyurea as demonstrated by the low rates of resistance, thrombosis, and disease progression. The main limitations of this study were the small patient cohort and the limited samples from PV diagnosis.