This is the first in vivo study that shows, that vorinostat has the potential to sensitize MRT to radiation.
Patients with malignant rhabdoid tumors still have a poor prognosis despite intensive current treatment protocols. Radiotherapy plays a significant role in local treatment, but is often delayed as long as possible in young children as late sequelae are feared. Therfore there is a need for new treatment strategies.
Epigenetic targeting therapies with HDACi have shown promising results in different tumor types. SAHA is one of the well-established HDACi which has been approved in the United States by the Food and Drug Administration (FDA) for the therapy for human cutaneous T-cell lymphoma. The drug can be given orally and side-effects are minor compared to cytotoxic chemotherapeutics. Therefore, it is an interesting candidate for tumor therapy. The promise of HDACi for cancer treatment, as a single agent and in combination with standard therapies, is supported by in vitro results of colorectal carcinoma, human melanoma and glioma cell experiments [4, 5, 7]. In several trials combination of HDACi with chemotherapy or radiotherapy improved tumor cell kill [4, 7, 16, 17]. However, in vivo studies with HDACi as single-agent in cancer treatment, showed only moderate and limited efficacy .
Based on our promising in vitro results that showed a selectively radiosensitizing effect in the A-204 cell line used in this experiment as well as in two osteosarcoma cell lines, we now investigated the efficacy of the HDACi SAHA in combination with radiotherapy, compared to SAHA or XRT only in a MRT xenograft mouse model.
The results of our first Cohort did show a trend towards an improved local control and tumor growth delay in mice treated with SAHA plus XRT compared to animals treated with XRT alone. We attributed the lack of significance in Cohort I to three possible parameters: 1. short period of SAHA application, 2. fractionation of radiation and 3. rather big tumor size at time of treatment initiation.
SAHA has previously been shown to augment the effects of radiotherapy in vivo tumor models [3, 6, 16]. In some of these in vivo studies, the drug was applied for several weeks with successful radiosensitization [7, 19], even though in others short-term application for 5 days and even single-dose treatment prior to irradiation seemed to be sufficient [7, 19].
The underlying pathways of HDACi acting as a radiosensitizer are still not completely understood . Abrogation of DSB repair has been suggested to be one possible mechanism by several authors [10, 16]. In our experiments, γH2AX expression, representing DNA-damage, was significantly influenced by the combination treatment. Interestingly, γH2AX expression of irradiated tumors reached the level of the untreated controls 24 h after XRT, whereas the tumors of the combination group showed still an elevated expression level (2-fold ± 0.5), thus showing an impairment of repair kinetics by the combination treatment. Therefore, we assumed improved results after fractionated XRT and decided to treat Cohort II with 3 × 3Gy instead of 1 × 10 Gy.
We wanted to use our first Cohort to investigate the potential of PET-imaging in our malignant rhabdoid tumor model. As the results were disappointing, we refrained from PET-imaging and thus were able to start treatment in smaller tumors in Cohort II.
These changes did indeed lead to a higher clinical impact resulting in significancy in Cohort II. As we decided to change all three parameters we can only guess which change has the highest impact and it remains unclear, if fractionation is obligatory or not. In previous reports on xenograft studies concerning breast tumors, neuroblastoma and ovarian cancer combination of HDACIs with fractionated [7, 17] as well as single-dose irradiation  have been reported to be successful.
Interestingly - in contrast to reports concerning in vivo models of other tumor types [16, 17, 20]- application of SAHA as a single agent had no effect in both of our experimental designs in this MRT xenotransplant model - neither when given for 8 days, nor for 3 weeks. Our SAHA dose, especially in Cohort II with 100 mg/kg for 5 consecutive days for 3 weeks is rather high compared to other studies [7, 17, 19, 20], in which SAHA proved to be successful. In these studies application differed from once 50 mg/kg, 3 × 150 mg within 1 week, 5 × 12,5 mg/week for 3 weeks and indeed 24 × 50 mg within 8 weeks. We therefore conclude from our study, that SAHA only treatment in MRT is not too promising.
We observed a strong trend towards a lower proliferation activity in the combination groups compared to the XRT alone groups. However, this trend failed to reach statistical significance in both cohorts.
Looking at necrosis of tumors, results showed no significant differences between the control group, SAHA group and the XRT or XRT plus SAHA group. However, apoptosis was indeed significantly higher in the XRT (p = 0,001) as well as XRT + SAHA groups (p = 0,0001) compared to the controls in both cohorts and also significantly higher in both XRT + SAHA groups compared to SAHA only (p > 0,0016). Thus, induction of apoptosis rather than necrosis seems to have the higher impact on radiosensitization by SAHA. However, as far as apoptosis was concerned the effect of SAHA + XRT seemed to be rather additive, while tumor growth delay in the mice seemed to be a rather synergistic effect given that SAHA alone had no significant effect at all. This proves again that apoptosis remains to be just one of the -all in all not completely understood - underlying mechanisms of radiosensitization.
We attribute the fact that no changes in our dynamic 18F-FDG-PET/CT images were to be observed to this lack of all investigated treatment schedules to induce significant necrosis in MRT. However, a further reason may be the initial tumor size (mean tumor size 300 mm3), which we chose in Cohort I in order to at all allow PET imaging , but which also may have attributed to the lower impact of the overall treatment that was observed in Cohort I compared to Cohort II.