Tamoxifen Reduces Mutation Allele Burden in MPN Subset Identified Via RNA Sequencing

Simón Méndez-Ferrer, PhD

The phase 2 TAMARIN study exploring the activity of tamoxifen on driver variant allele frequency (VAF) in patients with stable myeloproliferative neoplasms (MPN) met its primary end point, said Simón Méndez-Ferrer, PhD, who presented the data during the virtual 2020 ASH Annual Meeting & Exposition.

RNA sequencing in the study was found to predict which patients respond to tamoxifen—those with high levels of STAT3 and STAT5 activation as well as possessing an inflammation signature.

Of 32 patients who completed 24 weeks of tamoxifen treatment, 3 responders showed at least a 50% reduction in VAF, which was the study’s primary end point. A further 5 patients were classified as minor responders, showing reductions of at least 25% but lower than 50% in allele burden. One apparent responder had to be excluded from the analysis because the sample demonstrating at least a 50% reduction in allele burden was not confirmed.

“We found that tamoxifen reduces allele burden in a subset of MPN patients who could potentially be prospectively identified based on their transcriptome at baseline,” said Méndez-Ferrer, from the Department of Hematology, in the University of Cambridge, United Kingdom.

TAMARIN is a multicenter, single-arm, A’herns design, phase 2 clinical trial to assess the safety and effect of tamoxifen on molecular markers of disease in well-managed patients with MPN.

Males older than 60 years and postmenopausal women with stable blood counts and at least 20% mutated JAK2^V617F, CALRd^el52 or CALR^ins5 in the peripheral blood were eligible. Patients were excluded if they received treatment with alpha interferon or other new agents for their MPN within 6 months of trial entry or had a previous thrombotic event at any time.

Tamoxifen was given at 20 mg/day, and was escalated to 40 mg at 12 weeks if allele burden was not improved or a hematologic response was not attained.

Thirty-eight patients (27 male, 11 female) were treated; their mean age at baseline was 66.3 years. The time since diagnosis to study entry was a mean of 7.1 years. The MPN subtypes were essential thrombocythemia (n = 14), primary myelofibrosis (n = 6), post-essential thrombocythemia myelofibrosis (n = 2), polycythemia vera (n = 11), and post-polycythemia vera myelofibrosis (n = 5). About three-fourths (78.9%) had received 1 therapy prior to tamoxifen, mostly hydroxycarbamide (71.1%), which they continued to receive during the trial.

Thirty-two patients completed 24 weeks of tamoxifen treatment, and 31 patients continued on treatment after 24 weeks. “Two thrombotic events occurred in nonresponders, with 1 deep vein thrombosis after 20 weeks of treatment and 1 superficial thrombophlebitis after discontinuation in a patient who had already been 28 weeks on study,” said Méndez-Ferrer.

RNA sequencing from peripheral blood hematopoietic progenitor cells was performed at various time points. “Transcriptomic analysis segregated responders from nonresponders perfectly at baseline,” he said. “Gene expression enrichment analysis showed higher activation of hematopoietic progenitors in the responders, with increased expression of STAT3, STAT5, and inflammation-related genes. Additionally, hematopoietic progenitors from responders showed signs of integrative stress response manifested as increased expression of genes responsive to starvation or related to reactive oxygen species and apoptosis.”

Human cell lines (HEL and UKE-1) carrying the JAK2V617F mutation demonstrated dose-dependent apoptosis on HEL cells, confirming a direct apoptotic effect of tamoxifen. UKE-1 cells were sensitive to tamoxifen only under serum starvation, providing a model of sensitive and resistant cells with the same genetic background.

Gene expression enrichment analysis also showed increased expression of oxidative phosphorylation-related genes in the responders compared with nonresponders at baseline, and oxidative phosphorylation-related genes were significantly downregulated in responders only after tamoxifen treatment.

Analysis by RNA sequencing showed that tamoxifen downregulated oxidative phosphorylated genes, which was most pronounced in the sensitive cell lines, confirming the results obtained with the primary cells.

Mitochondrial restoration was measured in JAK2V617-mutated cell lines treated with tamoxifen or vehicle. Tamoxifen resulted in pronounced mitochondrial restoration and ATP production in sensitive JAK2-mutant cell lines.

Momethyl succinate was added in culture and was able to prevent tamoxifen-induced apoptosis in HEL cells. This protective effect of tamoxifen by momethyl succinate was explained by the rescue of mitochondrial restoration and ATP production. “It has been shown previously that ATP binding to the pseudokinase domain of JAK2 is critical for pathogenic activation of the mutant JAK2 but not for the wild-type JAK2,” said Méndez-Ferrer.

Further experiments suggest that tamoxifen may reduce mitochondrial activity and ATP generation to a critical level that might impair mutant JAK2 activity including its capacity to protect mutant cells from normal apoptosis.

Reference

Harrison C, Baxter J, Boucher RH, et al. Effects of tamoxifen on the mutant allele burden and disease course in patients with myeloproliferative neoplasms – results of the Tamarin Study. Presented at: 2020 ASH Annual Meeting and Exposition; December 4-8, 2020. Abstract 342.