CHOP and CHOP-like chemotherapy remain the most commonly used regimens for the treatment of peripheral T-cell lymphomas (PTCLs) despite sub-optimal results. Histone deacetylase inhibitors (HDACIs) are presently approved for the treatment of relapsed or refractory cutaneous T- cell lymphomas (CTCL) and peripheral T-cell lymphomas (PTCL) given their marked single agent activity in these diseases. The interaction between the HDACIs (depsipeptide (R) and belinostat (B)) and a DNMT inhibitor (decitabine (D)) was investigated in vitro, in vivo and at the molecular level in different T-cell lymphoma and leukemia cell lines including CTCL (H9, HH), and T- acute lymphoblastic leukemia (T-ALL) lines resistant to gamma-secretase inhibitors (P12, PF-382). For all cytotoxicity assays, a luminescence based cell viability assay was used (CellTiter-Glo™) followed by acquisition on a Biotek Synergy HT. Drug: drug interactions were analyzed using the calculation of the relative risk ratios (RRR<1 are defining synergism). Apoptosis was assessed by staining with Yo-Pro-1 and propidium iodine followed by FACSCalibur acquisition and analyzed using FlowJo. The IC50s for B, R, vorinostat (V), panobinostat (P), D and 5-Azacytidine alone were assessed at 24, 48 and 72 hours in all the cell lines. For the combination experiment we selected the most active DNMTI, decitabine. In the cytotoxicity assays, the combination of D plus B, R, V or P at 72 hours showed synergism in all the cell lines studied. The RRRs for all the combinations were between 0.0007 and 0.9. When H9, HH, P12 and PF382 cell lines were treated with D and B or R for 72 hours, all the combination groups showed significantly more apoptosis than the single drug exposures and controls. Table 1 displays the range of apoptosis induction for B, R ± D and the RRR value for the most significant data.
An
in vivo xenograft study in 6–8 weeks old female SCID beige mice injected subcutaneously with 2 × 10
7 HH cells was performed. Mice were separated into different cohorts and treated i.p. for 3 cycles with D or B or their combination according to the following schedules: D at 1.5 mg/kg on days 1, 3, 5; B at 40 mg/kg/day for 10 days (I cycle); D at 1.5mg/kg on days 15,17,19,21; B at 65 mg/kg/day for 10 days (II cycle); D at 1.5 mg/kg on days 29,31,33,35,37,39,41,43; B at 100mg/kg for 19 days (III cycle). Statistically significantly tumor growth inhibition was observed in the combination cohort compared to all the other cohorts (analysis on day 42, 45). We analyzed the molecular basis for this synergistic effect by evaluating gene expression patterns using the Illumina Human HT-12 v4 Expression BeadChip microarrays. These analyses revealed differentially expressed genes and modulated pathways for each of the single treatment conditions and the combination. As shown in
Figure 1, a set of genes (A) is down-regulated by both drugs. Other genes (B) are up-regulated by D and the effect is maintained in the combination. Other genes (C+E) are slightly up-regulated by R, though not significantly modified by D, and more strongly up-regulated in the combination group. Similarly, genes to some extent up-regulated by D but not by R (D+F) appeared to be more significantly affected by the combination. As shown in
Figure 2, the effects of the two drugs are largely different (only 39 genes modified in common by all the treatment groups). Most of the effects induced by the single agent treatment are maintained in the combination group (174 genes out of 191 for romidepsin and 211 genes out of 221 for decitabine). Interestingly, an additional 944 genes appeared to be modulated uniquely by the combination treatment strongly supporting the hypothesis of synergism also at the molecular level. Collectively, the data suggest that the combination of a DNMTI and HDACIs is synergistic in
in vitro and
in vivo model of T-cell lymphoma and is able to synergistically reverse the malignant signature at the molecular level. These data may constitute the basis for future phase I-II clinical trials.
