**4. Conclusions**

#### **4.1 Glioblastoma cells express a diverse radiosensitivity phenotype**

Our studies show that cell lines designated as "glioblastoma" in the literature are diverse in their radiosensitivity phenotypes.

#### **4.1.1 Some glioblastoma cell lines express a low rate of inactivation over the alpha\* response**

In our studies these reduced rates underlay the observation that such cell lines are refractory to doses over the alpha response. Specifically we hypothesized that these cell lines, express

Radiobiology of Radioresistant Glioblastoma 19

The data in figures 1, 2 and 3 show that U87 cells that express the VR phenotype but also express wtTP53 are more sensitive to doses that elicit the omega response (> 3 Gy) than the other two VR cells. This would suggest that tumors comprised of this form of VR cells would show significantly more inactivation than VR cells that express mut TP53 when doses

**5.1.3 The VR cell line U 251 shows increased radiosensitivity of its xenograft tumors** 

This increased response in vivo is observed for all protocols for glioblastoma cell line U-251 but is significantly elevated for protocols that use large fractions of 7.5 Gy and 15.0 Gy. This in turn suggests that some forms of VR glioblastoma cells will respond to hypofractionation

**6. Our observations suggest certain studies are needed to design protocols to exploit the VR radiosensitivity phenotype observed in some glioblastoma** 

Our research suggests several studies would improve selection of radiotherapy protocols

**6.1 Better markers are needed to define the VR radiosensitivity phenotype from tumor** 

Our studies overall suggest there is no single protocol that would be predicted to provide maximum improvement in tumor radiotherapy for all variations in the radiosensitivity

It seems clear from our work and that of others that a larger number of presumed glioblastoma cells need to be examined for their radiosensitivity phenotype including response to high dose-rate, to low dose-rate and their response as xenograft tumors to

**6.3 The mechanisms that underlay increased in vivo response of tumors comprised of** 

Our data show this is a significant increase in radiosensitivity and the mechanisms that underlay it need to be studied in detail. The data we have presented offers a useful range of responses in different genotypes to study the role of genotype, fraction-size and total dose

Radiosensitivity phenotypes of tumor cells are comprised of distinct, multiple responses to radiation. Glioblastoma cells exhibit responses that are both sensitive and resistant

phenotypes that we have observed in cells believed to be "glioblastoma" cells.

**U-251 glioblastoma cells needs to be extended to other glioblastoma cells** 

**6.2 The data base on the VR radiosensitivity phenotype need to be extended to** 

**5.1.2 VR glioblastoma cells that express wtTP53 should be more responsive to** 

**protocols that use higher doses per fraction (> 3 Gy)** 

are used that elicit the omega\* response.

with fractions circa 7.5 to 15.0 Gy.

**cell lines** 

**biopsies** 

selected protocols.

on this effect.

**7. Overall conclusions** 

**irradiated in vivo compared to other cell types** 

that could improve therapeutic results from specific protocols.

**include more cells presumed to be glioblastoma cells** 

"glio", an unidentified genetic or epigenetic factor, that renders such glioblastoma cells resistant to radiation delivered at doses below circa 3 Gy. This in turn suggests that tumors comprised of cells that express glio and fall into the VR radiosensitivity group will be refractory to radiotherapy that is based on multi-fraction of cells when fractions are below 3.0 Gy.

### **4.1.2 The alpha\* response can be uncoupled from the omega response in VR resistant glioblastoma cells**

Our data show that while two radioresistant glioblastoma cell lines (U-251 and T98G) show resistance to inactivation at higher doses (omega response), one cell line (U-87) shows a more sensitive response. We hypothesize that this difference is associated with expression of TP53 in these cell lines, U-251and T98G express mutTP53 while U-87 expresses wtTP53. These differences are consistent with the ration between alpha and omega responses for cells that express mutTP53 versus wtTP53 as shown in figure 3 and are consistent with all VR cells expressing glio but susceptible to the effect of expression of TP53.

#### **4.1.3 Glioblastoma cells that express VR radiosensitivity after high dose-rate irradiation show an unpredicted sensitivity to low dose-rate irradiation for dose-rates circa 0.25 to 0.49 Gy/hr**

Glioblastoma cell lines that show VR radiosensitivity secondary to their expression of glio when irradiated with high dose-rate show a relatively more sensitive response to low doserate irradiation for dose-rates circa 0.25 to 0.49 Gy/hr. The data in figures 5 and 6 show that two VR lines, U-87 and U-251 show rates of inactivation when irradiated with lower doserates that are more sensitive than the rates of inactivation over the alpha response igh doserate irradiation, response to LDR irradiation by exhibiting a more sensitive response show elevated rates of inactivation.

#### **4.1.4 Protocols that combine brachytherapy with external beam are highly effective in treating xenograft tumors of glioblastoma cells**

Led by JA Williams (27) we have shown that xenograft tumors of U-251 glioblastoma cells are highly susceptible to the combined effects of brachytherapy using single seeds and external beam radiotherapy. These data shows that xenograft tumors that are resistant to lower fractions (2 and 5 Gy) and protracted radiation from implanted radioactive seeds are highly response to their response to protocols that combine these two modalities.

#### **5.1 Implications for new approaches to radiotherapy of glioblastoma cell lines based on their radiosensitivity phenotype**

Our data suggest relationships between the radiosensitivity phenotypes of glioblastoma cells and their predicted response to different radiotherapy protocols.

#### **5.1.1 Cells that express the VR radiosensitivity phenotype will be refractory to protocols that use multiple fractions of doses lower than 3.0 Gy**

 Cells the express the VR radiosensitivity phenotype have low rates of inactivation by doses below 3 Gy (alpha response) and should not respond well to protocols that use smaller fraction sizes.

"glio", an unidentified genetic or epigenetic factor, that renders such glioblastoma cells resistant to radiation delivered at doses below circa 3 Gy. This in turn suggests that tumors comprised of cells that express glio and fall into the VR radiosensitivity group will be refractory to radiotherapy that is based on multi-fraction of cells when fractions are below

**4.1.2 The alpha\* response can be uncoupled from the omega response in VR resistant** 

Our data show that while two radioresistant glioblastoma cell lines (U-251 and T98G) show resistance to inactivation at higher doses (omega response), one cell line (U-87) shows a more sensitive response. We hypothesize that this difference is associated with expression of TP53 in these cell lines, U-251and T98G express mutTP53 while U-87 expresses wtTP53. These differences are consistent with the ration between alpha and omega responses for cells that express mutTP53 versus wtTP53 as shown in figure 3 and are consistent with all

VR cells expressing glio but susceptible to the effect of expression of TP53.

**4.1.3 Glioblastoma cells that express VR radiosensitivity after high dose-rate** 

**irradiation show an unpredicted sensitivity to low dose-rate irradiation for dose-rates** 

Glioblastoma cell lines that show VR radiosensitivity secondary to their expression of glio when irradiated with high dose-rate show a relatively more sensitive response to low doserate irradiation for dose-rates circa 0.25 to 0.49 Gy/hr. The data in figures 5 and 6 show that two VR lines, U-87 and U-251 show rates of inactivation when irradiated with lower doserates that are more sensitive than the rates of inactivation over the alpha response igh doserate irradiation, response to LDR irradiation by exhibiting a more sensitive response show

**4.1.4 Protocols that combine brachytherapy with external beam are highly effective in** 

Led by JA Williams (27) we have shown that xenograft tumors of U-251 glioblastoma cells are highly susceptible to the combined effects of brachytherapy using single seeds and external beam radiotherapy. These data shows that xenograft tumors that are resistant to lower fractions (2 and 5 Gy) and protracted radiation from implanted radioactive seeds are

**5.1 Implications for new approaches to radiotherapy of glioblastoma cell lines based** 

Our data suggest relationships between the radiosensitivity phenotypes of glioblastoma

 Cells the express the VR radiosensitivity phenotype have low rates of inactivation by doses below 3 Gy (alpha response) and should not respond well to protocols that use smaller

highly response to their response to protocols that combine these two modalities.

**5.1.1 Cells that express the VR radiosensitivity phenotype will be refractory to** 

cells and their predicted response to different radiotherapy protocols.

**protocols that use multiple fractions of doses lower than 3.0 Gy** 

3.0 Gy.

**glioblastoma cells** 

**circa 0.25 to 0.49 Gy/hr** 

elevated rates of inactivation.

**on their radiosensitivity phenotype** 

fraction sizes.

**treating xenograft tumors of glioblastoma cells** 

#### **5.1.2 VR glioblastoma cells that express wtTP53 should be more responsive to protocols that use higher doses per fraction (> 3 Gy)**

The data in figures 1, 2 and 3 show that U87 cells that express the VR phenotype but also express wtTP53 are more sensitive to doses that elicit the omega response (> 3 Gy) than the other two VR cells. This would suggest that tumors comprised of this form of VR cells would show significantly more inactivation than VR cells that express mut TP53 when doses are used that elicit the omega\* response.

#### **5.1.3 The VR cell line U 251 shows increased radiosensitivity of its xenograft tumors irradiated in vivo compared to other cell types**

This increased response in vivo is observed for all protocols for glioblastoma cell line U-251 but is significantly elevated for protocols that use large fractions of 7.5 Gy and 15.0 Gy. This in turn suggests that some forms of VR glioblastoma cells will respond to hypofractionation with fractions circa 7.5 to 15.0 Gy.
