Addition of the mTOR Inhibitor Rapamycin Deepens Response to HDAC Inhibition or Hypomethylating Therapy in Myelodysplastic Syndromes

Abstract 3824

Myelodysplastic syndromes (MDS) are clonal stem cell disorders characterised by ineffective blood cell production and increased leukaemia risk. Disordered regulation of the epigenome has been implicated in the pathogenesis of MDS and treatment with the hypomethylating agent azacitidine has a proven survival benefit in MDS cohorts. Other novel agents under investigation for the treatment of MDS include histone deacetylase inhibitors (HDACi) and small molecule inhibitors that target the PI3K/AKT/mTOR signal transduction pathway. PI3K/AKT/mTOR pathway inhibitors modify cellular translational capacity and translational efficiency providing a complementary mechanism of action to agents that act through epigenetic modification such as hypomethylating agents and HDACi.

We hypothesized that addition of an mTOR inhibitor (mTORi) would enhance the sensitivity of MDS clones to therapy with HDACi or hypomethylating agents and that responses would be karyotype dependent. Accordingly we sought to stratify in vitro responses to treatment with rapamycin (mTORi) alone, sodium valproate (HDACi) alone, decitabine (hypomethylating agent) alone, combination rapamycin plus sodium valproate or combination rapamycin plus decitabine.

Residual bone marrow from patients with a new diagnosis of MDS was salvaged after completion of the diagnostic work-up (n=18). Patients had a diagnosis of RA (n=2), RARS (n=2), RCMD (n=11) and RAEB-2 (n=3). 7/18 (39%) patients had a karyotypic abnormality detected by metaphase cytogenetics. IPSS-R cytogenetic risk categories represented were: very good (n=3), good (n=11), intermediate (n=2) and poor (n=2). Genomic DNA was hybridized to a 300K single nucleotide polymorphism array (SNP-A). A karyotype abnormality score was assigned for the total number of independent cytogenetic abnormalities detected by combined use of metaphase cytogenetics and SNP-A. Bone marrow mononuclear cells were plated in duplicate in methylcellulose containing SCF, GM-CSF, IL-3 and EPO together with vehicle control, 1mM sodium valproate, 100nM rapamycin, 300uM decitabine, 1mM sodium valproate plus 100nM rapamycin or 300uM decitabine plus 100nM rapamycin. Colony formation was scored on day 14. Drug responses were classified as sensitive (colony numbers <25% of vehicle control plate), resistant (colony numbers ≥75% of control) or intermediate (neither sensitive nor resistant).

SNP-A detected cytogenetically cryptic abnormalities in 6/18 (33%) cases. The karyotype abnormality score ranged between 0 and 3. Suppression of colony formation in response to single agent therapy was variable. The number of cases deemed sensitive, intermediate and resistant respectively for rapamycin was 3, 9 and 6. For sodium valproate it was 2, 8 and 8 and for decitabine 3, 11 and 4. Response to rapamycin combination therapies was higher than response to any single agent. The number of cases deemed sensitive, intermediate and resistant was 9, 8 and 1 respectively for combination rapamycin and sodium valproate and the corresponding numbers for combination rapamycin and decitabine were 8, 8, and 2. Response to rapamycin was more likely in patient samples with less karyotypic complexity (average colony growth 41.4±9.9% control for cases with a karyotype score of 0–1 vs 76.1±12.0% for cases with a score of 2–3, p<0.05). The same was true of combination rapamycin and decitabine (average colony growth 19.5±4.9% control for cases with a karyotype score of 0–1 vs 54.5±12.5% for cases with a score of 2–3, p=0.01). The response rate to sodium valproate was higher in samples from patients with isolated erythroid dysplasia (average colony growth 29.6±6.0% control for unilineage dysplasia versus 92.2±12.7% for trilineage dysplasia, p=0.02). Best response (defined as the greatest suppression of colony growth in each patient sample) was observed after combined rapamycin and sodium valproate in 9 patients, combined rapamycin and decitabine in 8 patients and rapamycin alone in 1 patient. Patients with higher risk disease (2 or more karyotypic abnormalities and/or an increased number of bone marrow blasts) were more likely to respond to combination rapamycin and sodium valproate therapy (4/8 (50%) response rate vs 2/8 (25%) for combination rapamycin and decitabine). This is evidence of benefit from the rational addition of mTOR inhibitor therapy to treatment with hypomethylating agents or HDACi in a preclinical model of MDS.