**Ultrafractionated Radiation Therapy (3 Daily Doses of 0.75 Gy) - A New and Promising Radiotherapy Schedule for Glioblastoma Patients**

Patrick Beauchesne *Neuro-Oncology Department, CHU de NANCY France* 

#### **1. Introduction**

396 Advances in the Biology, Imaging and Therapies for Glioblastoma

[17] Leong, R., Nguyen, N., Meredith, C., Al-Sohaily, S., Kukic, D., Delaney, P., Murr, E.,

diagnosis and evaluation of celiac disease. *Gastroenterology,* 135, 1870-1876 [18] Günther, U., Daum, S., Heller, F., Schumann, M., Loddenkemper, C., Grünbaum, M.,

[19] Louis, D. N., Ohgaki, H., Wiestler, O. D., Cavenee, W. K., Burger, P. C., Jouvet, A.,

[20] Wessels, J. T., Busse, A. C., Mahrt, J., Dullin, C., Grabbe, E. & Mueller, G. A. (2007). In

[21] Sanai, N., Eschbacher, J., Hattendorf , G., Coons, S. W., Preul, M. C., Smith, K. A.,

[23] Lacroix, M., Abi-Said, D., Fourney, D. R., Gokaslan, Z. L., Shi, W., DeMonte, F., Lang, F.

Brain Tumors: A Feasibility Analysis in Humans. *Neurosurgery,* Epub, [22] Schlosser, H.G., Bojarski, C. (2011 (Epub)). Confocal Neurolasermicroscopy (NLM).

glioblastomas and anaplastic astrocytomas. *Neurosurgery,* 21, 2, 201-6 [25] Burger, P. C. & Green, S. B. (1987). Patient age, histologic features, and length of survival in patients with glioblastoma multiforme. *Cancer,* 59, 9, 1617-25 [26] Wood, J. R., Green, S. B. & Shapiro, W. R. (1988). The prognostic importance of tumor

Cooperative Group. *J Clin Oncol,* 6, 2, 338-43

injections. *Int Ophthalmol Clin,* 29, 3, 200-5

the central nervous system. *Acta Neuropathol,* 114, 2, 97-109

celiac disease. *Endoscopy,* 42, 3, 197-202

8, 542-9

Neurosurgery, Epub

11, 4, 254-67

Yong, J., Merrett, N. & Biankin, A. (2008). In vivo confocal endomicroscopy in the

Zeitz, M. & Bojarski, C. (2010). Diagnostic value of confocal endomicroscopy in

Scheithauer, B. W. & Kleihues, P. (2007). The 2007 WHO classification of tumours of

vivo imaging in experimental preclinical tumor research--a review. *Cytometry A,* 71,

Nakaji, P. & Spetzler, R. F. (2011 (Epub)). Intraoperative Confocal Microscopy for

F., McCutcheon, I. E., Hassenbusch, S. J., Holland, E., Hess, K., Michael, C., Miller, D. & Sawaya, R. (2001). A multivariate analysis of 416 patients with glioblastoma multiforme: prognosis, extent of resection, and survival. *J Neurosurg,* 95, 2, 190-8 [24] Ammirati, M., Vick, N., Liao, Y. L., Ciric, I. & Mikhael, M. (1987). Effect of the extent of

surgical resection on survival and quality of life in patients with supratentorial

size in malignant gliomas: a computed tomographic scan study by the Brain Tumor

S., Knuttel, A. & Sterry, W. (2007). Application of optical non-invasive methods in skin physiology: a comparison of laser scanning microscopy and optical coherent

confocal microscopy of skin in vivo: microscope and fluorophores. *Skin Res Technol,* 

Coons, S. W., Scheck, A. C., Smith, K. A., Spetzler, R. F. & Preul, M. C. (2010). Miniaturized handheld confocal microscopy for neurosurgery: results in an

[27] Asthagiri, A. R., Pouratian, N., Sherman, J., Ahmed, G. & Shaffrey, M. E. (2007). Advances in brain tumor surgery. *Neurol Clin,* 25, 4, 975-1003, viii-ix [28] Lademann, J., Otberg, N., Richter, H., Meyer, L., Audring, H., Teichmann, A., Thomas,

[30] Sankar, T., Delaney, P. M., Ryan, R. W., Eschbacher, J., Abdelwahab, M., Nakaji, P.,

[31] Makale, M. (2007). Intravital imaging and cell invasion. *Methods Enzymol,* 426, 375-401 [32] Lipson, B. K. & Yannuzzi, L. A. (1989). Complications of intravenous fluorescein

tomography with histological analysis. *Skin Res Technol,* 13, 2, 119-32 [29] Suihko, C., Swindle, L. D., Thomas, S. G. & Serup, J. (2005). Fluorescence fibre-optic

experimental glioblastoma model. *Neurosurgery,* 66, 2, 410-417

Malignant glioma is one of the most radio-resistant tumor types and accounts for approximately 60% of all primary brain tumors in adults (Behin et al., 2003; Black, 1991a, 1991b; DeAngelis, 2001). There are three distinct histological types: anaplastic astrocytoma (AA), anaplastic oligodendroglioma (AO), and glioblastoma multiforme (GBM). The prognosis of malignant glioma patients remains dismal (Behin et al., 2003; Black, 1991a, 1991b; De Angelis, 2001). The median survival for patients with newly diagnosed GBM is 8 to 15 months, prognosis is slightly better for newly diagnosed AA with a median survival of 24 to 36 months, and the prognosis for AO gives a median survival of 60 months (Behin et al., 2003; Black, 1991a, 1991b; De Angelis, 2001). For AA and GBM, the standard of care consists of surgical resection of as much of the tumor as is considered to be safe, followed by radiation and chemotherapy and has been so for many decades (Behin et al., 2003; Black, 1991a, 1991b; DeAngelis, 2001, Fine et al., 1993; Stewart, 2002; Walker et al., 1978, 1980). A new standard procedure for GBM has recently been defined by the EORTC phase III trial which randomized patients in two groups, receiving either temozolomide (TMZ) concomitant and adjuvant to radiation therapy or radiation therapy alone (Stupp et al., 2005). A significant increase in overall survival (OS) was seen in the radiation therapy plus TMZ group compared to the radiation therapy alone group. Survival rates were respectively 14.6 and 12.1 months. For AO, the standard treatment is surgical resection followed by radiation therapy (Stupp et al., 2005). Adjuvant chemotherapy does not provide significant benefits in OS (Van den Bent et al., 2006).

Radiation therapy remains the backbone of care for glioblastomas, even in patients who have undergone a prior presumed complete resection. The infiltrative nature of these tumors makes a truly complete resection nearly impossible in most cases (Behin et al., 2003; Black, 1991a, 1991b; DeAngelis, 2001, Fine et al., 1993; Hall, 1978; Stewart, 2002; Walker et al., 1978, 1980). Standard fractionated radiation therapy delivers a total radiation dose of 60 Gy given in 30 fractions over 6 weeks. The target is usually the tumor bulk as visualized on CT or MRI, with a wide margin of 2-3 cm (Behin et al., 2003; Black, 1991a, 1991b; DeAngelis,

Ultrafractionated Radiation Therapy (3 daily doses of 0.75 Gy) -

Marples et al., 1994).

A New and Promising Radiotherapy Schedule for Glioblastoma Patients 399

phenomenon was defined and termed "low-dose hyper-radiosensitivity" (HRS), and the second phenomenon "increased radio-resistance" (IRR) (Joiner et al., 1986; Lambin et al., 1994b, 1996; Marples et al., 1997; Short et al., 1999b, 2001; Turesson & Joiner, 1996; Wouters et al., 1996). While the LQ model underestimates the HRS phenomenon, it correlates to the data at doses ranging from 2 – 5 Gy. HRS was represented as an undeniable downward "kink" on survival curve for doses below 1 Gy (Fig. 1). This was demonstrated by Wouters et al using the flow cytometry survival (FACS) method thus showing that it was not merely an artifact associated with the DMIPS assay (Wouters & Skarsgard, 1994). HRS has also been triggered in the human lung epithelial cell line, L132, after exposition to very low-doses of X-rays (Singh et al., 19974), and found with Chinese Hamster cells (Joiner et al., 1993a;

Fig. 1. Low-dose hypersensitivity was represented as an undeniable downward "kink" on survival curve for doses below 1 Gy, followed for doses superior to 2 Gy by "IRR" or

Lambin et al irradiated the HT 29 cell line, derived from a human colorectal tumor and considered as a radio-resistant tumor at usual X-ray doses, with single-doses of X-rays from 0.05 to 5 Gy. They focused on cell survival at doses of less than 1 Gy, using the DMIPS cell analyzer (Lambin et al., 1993a, 1993b). At doses < 0.5 Gy, an increased X-ray sensitivity was

"increased radio-resistance" phenomenon.

2001, Fine et al., 1993; Hall, 1978; Stewart, 2002; Walker et al., 1978, 1980). Although radiation therapy is not a curative treatment for glioblastomas, it results in prolongation of life with optimized quality (Behin et al., 2003; Black, 1991a, 1991b; DeAngelis, 2001, Fine et al., 1993; Stewart, 2002; Walker et al., 1978, 1980). Whether the clinical radioresistance of GBM is due solely to inherent radioresistance at the cellular level is unclear. Overall, malignant glioma cell lines exhibit SF2 (SF - 2 Gy) values at the upper end of the range compared with other human tumor cell lines, though studies have failed to link clinical response with SF2 (Taghian, 1992, 1993). Defining the molecular basis of radioresistance is, therefore, important. Disruption of cell-cycle arrest or apoptotic pathways by *INK4a* loss or by *p53* mutations or inactivation (approx. 40–60% of malignant gliomas have *p53* mutations) associated with *CDK4* amplification or *Rb* loss may be significant factors in determining the response of these tumors to irradiation and treatment outcome (James & Olson, 1986; Kleihues & Ohgaki, 1999, Watanabe et al., 1996).

The radiation survival response of mammalian cells is more complicated than once believed. A few studies indicate that some human cell lines are sensitive to killing by low radiation doses (1 Gy). This has been termed *low-dose hyper-radiosensitivity* (HRS) (Joiner et al., 1986; Lambin et al., 1994b, 1996; Marples et al., 1997; Short et al., 1999b; Turesson & Joiner, 1996). This phenomenon is more apparent in radioresistant cell lines such as glioma cells, and is substantially underestimated by the linear-quadratic (LQ) model (Joiner et al., 1986; Lambin et al., 1994b, 1996; Marples et al., 1997; Short et al., 1999b, 2001; Turesson & Joiner, 1996; Wouters et al., 1996). It may reflect differential triggering or induction of repair mechanisms. Cells may be sensitive to low doses because repair mechanisms are not induced, whereas higher doses may cause enough damage to induce or trigger repair mechanisms and, therefore, exhibit increased radioresistance (Joiner et al., 1986; Lambin et al., 1994b, 1996; Marples et al., 1997; Short et al., 1999b, 2001; Turesson & Joiner, 1996; Wouters et al., 1996). Still, new modalities of radiation therapy are urgently needed.

#### **2.** *In Vitro* **studies**

#### **2.1 Cell lines experiments**

The GRAY laboratory first demonstrated an increased X-ray sensitivity in murine skin and kidney after very low doses per fraction (Joiner & Denekamp, 1986; Joiner et al., 1986; Joiner & Johns, 1988). They irradiated the V79 murine fibroblast cell line with 250kVp X-rays and measured cell survival with a Dynamic Microscopic Imaging Processing Scanner (DMIPS) cell analyzer (Joiner et al., 1993a; Marples et al., 1994). Briefly, 3000-5000 cells were plated into 25 cm2 tissue culture flasks and left to incubate for 4 to 6 hours at 37° C. The flasks were removed from the incubator halfway through the initial 4-6 hour incubation period, the medium removed, and the flasks were then immediately completely refilled with fresh medium before being sealed. Following irradiation, the DMIPS cell analyser was used to locate and record the positions of 300-400 isolated cells within 10 cm2 in the centre of each flask. After 6-7 days of incubation at 37°C, all the originally recorded cell locations were revisited to assay for colony formation using a criterion for survival of 50 cells or more per colony as determined by manual microscopic examination of each selected location in the flask (Marples et al., 1994). The results displayed an increased X-ray sensitivity (hypersensitivity) after very small doses (< 0.3 Gy), followed by an increase in survival after the doses increased from 0.3 – 1 Gy (Joiner et al., 1993a; Marples et al., 1994). The first

2001, Fine et al., 1993; Hall, 1978; Stewart, 2002; Walker et al., 1978, 1980). Although radiation therapy is not a curative treatment for glioblastomas, it results in prolongation of life with optimized quality (Behin et al., 2003; Black, 1991a, 1991b; DeAngelis, 2001, Fine et al., 1993; Stewart, 2002; Walker et al., 1978, 1980). Whether the clinical radioresistance of GBM is due solely to inherent radioresistance at the cellular level is unclear. Overall, malignant glioma cell lines exhibit SF2 (SF - 2 Gy) values at the upper end of the range compared with other human tumor cell lines, though studies have failed to link clinical response with SF2 (Taghian, 1992, 1993). Defining the molecular basis of radioresistance is, therefore, important. Disruption of cell-cycle arrest or apoptotic pathways by *INK4a* loss or by *p53* mutations or inactivation (approx. 40–60% of malignant gliomas have *p53* mutations) associated with *CDK4* amplification or *Rb* loss may be significant factors in determining the response of these tumors to irradiation and treatment outcome (James & Olson, 1986;

The radiation survival response of mammalian cells is more complicated than once believed. A few studies indicate that some human cell lines are sensitive to killing by low radiation doses (1 Gy). This has been termed *low-dose hyper-radiosensitivity* (HRS) (Joiner et al., 1986; Lambin et al., 1994b, 1996; Marples et al., 1997; Short et al., 1999b; Turesson & Joiner, 1996). This phenomenon is more apparent in radioresistant cell lines such as glioma cells, and is substantially underestimated by the linear-quadratic (LQ) model (Joiner et al., 1986; Lambin et al., 1994b, 1996; Marples et al., 1997; Short et al., 1999b, 2001; Turesson & Joiner, 1996; Wouters et al., 1996). It may reflect differential triggering or induction of repair mechanisms. Cells may be sensitive to low doses because repair mechanisms are not induced, whereas higher doses may cause enough damage to induce or trigger repair mechanisms and, therefore, exhibit increased radioresistance (Joiner et al., 1986; Lambin et al., 1994b, 1996; Marples et al., 1997; Short et al., 1999b, 2001; Turesson & Joiner, 1996; Wouters et al., 1996).

The GRAY laboratory first demonstrated an increased X-ray sensitivity in murine skin and kidney after very low doses per fraction (Joiner & Denekamp, 1986; Joiner et al., 1986; Joiner & Johns, 1988). They irradiated the V79 murine fibroblast cell line with 250kVp X-rays and measured cell survival with a Dynamic Microscopic Imaging Processing Scanner (DMIPS) cell analyzer (Joiner et al., 1993a; Marples et al., 1994). Briefly, 3000-5000 cells were plated into 25 cm2 tissue culture flasks and left to incubate for 4 to 6 hours at 37° C. The flasks were removed from the incubator halfway through the initial 4-6 hour incubation period, the medium removed, and the flasks were then immediately completely refilled with fresh medium before being sealed. Following irradiation, the DMIPS cell analyser was used to locate and record the positions of 300-400 isolated cells within 10 cm2 in the centre of each flask. After 6-7 days of incubation at 37°C, all the originally recorded cell locations were revisited to assay for colony formation using a criterion for survival of 50 cells or more per colony as determined by manual microscopic examination of each selected location in the flask (Marples et al., 1994). The results displayed an increased X-ray sensitivity (hypersensitivity) after very small doses (< 0.3 Gy), followed by an increase in survival after the doses increased from 0.3 – 1 Gy (Joiner et al., 1993a; Marples et al., 1994). The first

Kleihues & Ohgaki, 1999, Watanabe et al., 1996).

Still, new modalities of radiation therapy are urgently needed.

**2.** *In Vitro* **studies** 

**2.1 Cell lines experiments** 

phenomenon was defined and termed "low-dose hyper-radiosensitivity" (HRS), and the second phenomenon "increased radio-resistance" (IRR) (Joiner et al., 1986; Lambin et al., 1994b, 1996; Marples et al., 1997; Short et al., 1999b, 2001; Turesson & Joiner, 1996; Wouters et al., 1996). While the LQ model underestimates the HRS phenomenon, it correlates to the data at doses ranging from 2 – 5 Gy. HRS was represented as an undeniable downward "kink" on survival curve for doses below 1 Gy (Fig. 1). This was demonstrated by Wouters et al using the flow cytometry survival (FACS) method thus showing that it was not merely an artifact associated with the DMIPS assay (Wouters & Skarsgard, 1994). HRS has also been triggered in the human lung epithelial cell line, L132, after exposition to very low-doses of X-rays (Singh et al., 19974), and found with Chinese Hamster cells (Joiner et al., 1993a; Marples et al., 1994).

Fig. 1. Low-dose hypersensitivity was represented as an undeniable downward "kink" on survival curve for doses below 1 Gy, followed for doses superior to 2 Gy by "IRR" or "increased radio-resistance" phenomenon.

Lambin et al irradiated the HT 29 cell line, derived from a human colorectal tumor and considered as a radio-resistant tumor at usual X-ray doses, with single-doses of X-rays from 0.05 to 5 Gy. They focused on cell survival at doses of less than 1 Gy, using the DMIPS cell analyzer (Lambin et al., 1993a, 1993b). At doses < 0.5 Gy, an increased X-ray sensitivity was

Ultrafractionated Radiation Therapy (3 daily doses of 0.75 Gy) -

10

radiation therapy.

Pedeux et al., 2003).

et al., 2003; Pedeux et al., 2003).

**2.2 Repeated irradiations** 

100

SF %

A New and Promising Radiotherapy Schedule for Glioblastoma Patients 401

CL35 G5 G111 G142 G152

0 0,3 0,5 0,6 0,7 0,8 0,9 1 1,5 2 Doses Gy

Fig. 2. Survival of human glioma cells following irradiation. Cells were irradiated with 0–2 Gy. G111, G142 and G152 glioma cell lines display HRS at doses below 1 Gy. HRS was observed only with G5 cells, whereas its clone, CL35, displayed conventional sensitivity to

The authors demonstrated HRS in four human melanoma tumor cell lines, M4Be – A375P – MeWo – SKMe12, at doses below 1 Gy, and in the MRC5 human fibroblast cell strain (Beauchesne et al., 2003; Pedeux et al., 2003). HRS was not expressed in two radio-sensitive cell lines, H460 (from a lung cancer) and MCF7 (from breast cancer) (Beauchesne et al., 2003;

The same team (Beauchesne et al) also tested the chemotherapy combination etoposide – temozolomide concomitantly with low-doses fractions on the human glioblastoma cell lines, G5 – G142 – G152 (Beauchesne et al., 2003; Pedeux et al., 2003). Cells were incubated immediately after an ultrafractionated irradiation regimen with etoposide and temozolomide at determined doses for 24 hours. A marked radio-sensizitation effect was observed with the CL35 line, and an enhancement of the HRS phenomenon was reported for G142 and G152 (Beauchesne et al., 2003; Pedeux et al., 2003). Thus, the combination of chemotherapy and radiotherapy enhances the effects of the therapies, thus further improving the effect of repeated low-radiation doses on malignant glioma cells (Beauchesne

Short et al in another set of experiments using the T98G human cell line (derived from a human glioblastoma), tested low-doses irradiations given at < 0.5 Gy once or more daily

observed. The HT 29 cell line was also irradiated with neutrons [d(4)-Be], obtained by a Van de Graaff accelerator bombarding a thick beryllium target with 4 MeV deuterons, at a dose rate of 0-20 Gy/min, but no HRS was observed (Lambin et al., 1993a, 1993b). In another study, Lambin et al studied an RT 112 cell line derived from a bladder carcinoma (Lambin et al., 1994a, 1994b). At a survival fraction of 60 % at 2 Gy (SF 2 Gy) this tumor was considered to be as radio-resistant (Lambin et al., 1994a, 1994b). The cell line was irradiated with low doses X-ray, and the HRS phenomenon observed at doses of < 0.5 Gy using the DMIPS method (Lambin et al., 1994a, 1994b).

The cell lines Be 11 and MeWo derived from melanoma, SW 48 from a colorectal tumor, and HX 142 from a neuroblastoma were irradiated with low doses X-ray as described above (Lambin et al., 1996). The cell line Be 11 was considered as radio-resistant - the SF 2 Gy ranged from 60 to 70 % - but the cell lines MeWo, SW 48 and HX 142 were radio-sensitive tumors with an SF 2 Gy ranging from 3 to 29 % (Lambin et al., 1996). The response obtained for doses ranging from 2 – 5 Gy for all cell lines fit with the LQ model, but HRS at doses < 0.5 Gy was not observed for the cell lines MeWo, SW 48 and HX 142 (Lambin et al., 1996). This absence of HRS in radiosensitive cell lines could be explained by the decreased inducible response of these cell lines (Lambin et al., 1996).

Human glioblastoma is considered to be one of the most radio-resistant tumors. Short et al studied five human glioblastoma cell lines, T98G – A7 – U87MG – U138 – HGL21, and one cell line derived from an anaplastic astrocytoma, U373 (Short et al., 1999a, 1999b). All the cell lines were irradiated with low doses of X-ray (Short et al., 1999a, 1999b). Survival time was calculated for the T98G – A7 – U373T9 lines using the DMIPS method. Survival time for U87MG – U138 – HGL21 which are not suitable for the DMIPS methods, were obtained from the cell shorter (CS) protocol as modified by Wouters et al (Wouters et al., 1996). HRS was noted at very low doses of X-rays in all five of the human glioblastoma cell lines, and most markedly in the A7 – U138 – TG98 cell lines (Short et al., 1999a, 1999b). The grade III cell line, U373, did not express HRS, though no clear explanation for this was put forward; possibly limitations of the CS methods or because this cell line could expresses HRS at much lower doses than is technically possible to test (Short et al., 1999a, 1999b).

To date, the low-dose responses have been reported by several laboratories in more than 26 different human cell lines, and survival times obtained using both DMIPS and the colony assay formation (CFA) (Beauchesne et al., 2003; Joiner et al., 1986, 1993a, 1993b; Lambin et al., 1993a, 1993b, 1994a, 1994b, 1994c, 1996; Marples et al., 1994, 1997; Short et al., 1999a, 1999b, 2001; Singh et al., 1994; Turesson & Joiner, 1996; Wouters et al., 1994, 1996). These include cell lines from colorectal carcinoma, bladder carcinoma, melanoma, prostate carcinoma, cervical squamous carcinoma, lung adenocarcinoma, neuroblastoma, gliomas, one non-malignant lung epithelial line, and one primary human fibroblast line.

Beauchesne et al also studied the HRS phenomenon in a French laboratory, using a linear accelerator to deliver the daily the radiation therapy for hospitalized patients (Beauchesne et al., 2003; Pedeux et al., 2003). The following human malignant cell lines established in this laboratory and previously described were tested; G5 – CL35 (a clone derived from G5) – G111 – G142 – G152. Cell survival was calculated from the CFA technique. Three hours after plating, cells were exposed to X-rays delivered by a linear accelerator (X photons of 10 MeV, dose rate of 2.43 Gy/min) with the irradiator placed at a distance of 1 m from the target, and an irradiation field of 40 X 40 cm (Beauchesne et al., 2003; Pedeux et al., 2003). The irradiation doses ranged from 0.2 to 2 Gy. HRS was once more reported at doses lower than 1 Gy for the glioma cell lines G5-G111-G142-G152, though, CL35, a regular sub-clone of G5, failed to express this HRS phenomenon (Fig. 2) (Beauchesne et al., 2003; Pedeux et al., 2003).

observed. The HT 29 cell line was also irradiated with neutrons [d(4)-Be], obtained by a Van de Graaff accelerator bombarding a thick beryllium target with 4 MeV deuterons, at a dose rate of 0-20 Gy/min, but no HRS was observed (Lambin et al., 1993a, 1993b). In another study, Lambin et al studied an RT 112 cell line derived from a bladder carcinoma (Lambin et al., 1994a, 1994b). At a survival fraction of 60 % at 2 Gy (SF 2 Gy) this tumor was considered to be as radio-resistant (Lambin et al., 1994a, 1994b). The cell line was irradiated with low doses X-ray, and the HRS phenomenon observed at doses of < 0.5 Gy using the DMIPS

The cell lines Be 11 and MeWo derived from melanoma, SW 48 from a colorectal tumor, and HX 142 from a neuroblastoma were irradiated with low doses X-ray as described above (Lambin et al., 1996). The cell line Be 11 was considered as radio-resistant - the SF 2 Gy ranged from 60 to 70 % - but the cell lines MeWo, SW 48 and HX 142 were radio-sensitive tumors with an SF 2 Gy ranging from 3 to 29 % (Lambin et al., 1996). The response obtained for doses ranging from 2 – 5 Gy for all cell lines fit with the LQ model, but HRS at doses < 0.5 Gy was not observed for the cell lines MeWo, SW 48 and HX 142 (Lambin et al., 1996). This absence of HRS in radiosensitive cell lines could be explained by the decreased

Human glioblastoma is considered to be one of the most radio-resistant tumors. Short et al studied five human glioblastoma cell lines, T98G – A7 – U87MG – U138 – HGL21, and one cell line derived from an anaplastic astrocytoma, U373 (Short et al., 1999a, 1999b). All the cell lines were irradiated with low doses of X-ray (Short et al., 1999a, 1999b). Survival time was calculated for the T98G – A7 – U373T9 lines using the DMIPS method. Survival time for U87MG – U138 – HGL21 which are not suitable for the DMIPS methods, were obtained from the cell shorter (CS) protocol as modified by Wouters et al (Wouters et al., 1996). HRS was noted at very low doses of X-rays in all five of the human glioblastoma cell lines, and most markedly in the A7 – U138 – TG98 cell lines (Short et al., 1999a, 1999b). The grade III cell line, U373, did not express HRS, though no clear explanation for this was put forward; possibly limitations of the CS methods or because this cell line could expresses HRS at much

To date, the low-dose responses have been reported by several laboratories in more than 26 different human cell lines, and survival times obtained using both DMIPS and the colony assay formation (CFA) (Beauchesne et al., 2003; Joiner et al., 1986, 1993a, 1993b; Lambin et al., 1993a, 1993b, 1994a, 1994b, 1994c, 1996; Marples et al., 1994, 1997; Short et al., 1999a, 1999b, 2001; Singh et al., 1994; Turesson & Joiner, 1996; Wouters et al., 1994, 1996). These include cell lines from colorectal carcinoma, bladder carcinoma, melanoma, prostate carcinoma, cervical squamous carcinoma, lung adenocarcinoma, neuroblastoma, gliomas,

Beauchesne et al also studied the HRS phenomenon in a French laboratory, using a linear accelerator to deliver the daily the radiation therapy for hospitalized patients (Beauchesne et al., 2003; Pedeux et al., 2003). The following human malignant cell lines established in this laboratory and previously described were tested; G5 – CL35 (a clone derived from G5) – G111 – G142 – G152. Cell survival was calculated from the CFA technique. Three hours after plating, cells were exposed to X-rays delivered by a linear accelerator (X photons of 10 MeV, dose rate of 2.43 Gy/min) with the irradiator placed at a distance of 1 m from the target, and an irradiation field of 40 X 40 cm (Beauchesne et al., 2003; Pedeux et al., 2003). The irradiation doses ranged from 0.2 to 2 Gy. HRS was once more reported at doses lower than 1 Gy for the glioma cell lines G5-G111-G142-G152, though, CL35, a regular sub-clone of G5, failed to express this HRS phenomenon (Fig. 2) (Beauchesne et al., 2003; Pedeux et al., 2003).

method (Lambin et al., 1994a, 1994b).

inducible response of these cell lines (Lambin et al., 1996).

lower doses than is technically possible to test (Short et al., 1999a, 1999b).

one non-malignant lung epithelial line, and one primary human fibroblast line.

Fig. 2. Survival of human glioma cells following irradiation. Cells were irradiated with 0–2 Gy. G111, G142 and G152 glioma cell lines display HRS at doses below 1 Gy. HRS was observed only with G5 cells, whereas its clone, CL35, displayed conventional sensitivity to radiation therapy.

The authors demonstrated HRS in four human melanoma tumor cell lines, M4Be – A375P – MeWo – SKMe12, at doses below 1 Gy, and in the MRC5 human fibroblast cell strain (Beauchesne et al., 2003; Pedeux et al., 2003). HRS was not expressed in two radio-sensitive cell lines, H460 (from a lung cancer) and MCF7 (from breast cancer) (Beauchesne et al., 2003; Pedeux et al., 2003).

The same team (Beauchesne et al) also tested the chemotherapy combination etoposide – temozolomide concomitantly with low-doses fractions on the human glioblastoma cell lines, G5 – G142 – G152 (Beauchesne et al., 2003; Pedeux et al., 2003). Cells were incubated immediately after an ultrafractionated irradiation regimen with etoposide and temozolomide at determined doses for 24 hours. A marked radio-sensizitation effect was observed with the CL35 line, and an enhancement of the HRS phenomenon was reported for G142 and G152 (Beauchesne et al., 2003; Pedeux et al., 2003). Thus, the combination of chemotherapy and radiotherapy enhances the effects of the therapies, thus further improving the effect of repeated low-radiation doses on malignant glioma cells (Beauchesne et al., 2003; Pedeux et al., 2003).

#### **2.2 Repeated irradiations**

Short et al in another set of experiments using the T98G human cell line (derived from a human glioblastoma), tested low-doses irradiations given at < 0.5 Gy once or more daily

Ultrafractionated Radiation Therapy (3 daily doses of 0.75 Gy) -

**3.** *In Vivo* **experiments** 

A New and Promising Radiotherapy Schedule for Glioblastoma Patients 403

2 days. Again, a marked and significant increase in cell killing occurred after the repeated low-doses for G5 but not the CL35 cell line (Beauchesne et al., 2003; Pedeux et al., 2003). It was postulated that te HRS phenomenon was responsible for the lower cell survival obtained after ultrafractionated regimen (Beauchesne et al., 2003; Pedeux et al., 2003).

The first study which tested ultrafractionated irradiation on an animal model was reported by Beck-Bornholdt; the rat rhabdomyosarcoma R1H was irradiated with 126 fractions over 6 weeks. Top-up irradiations were not given (different doses per fraction between 0.43 and 0.71 Gy were applied) (Beck-Bornholdt et al., 1989). The results were compared to "historical control", and the authors demonstrated that the ultrafractionated regimen was slight more

With a view to demonstrating a potential therapeutic benefit of the ultrafractionated irradiation schedule for malignant glioma patients, Beauchesne et al tested the fractionated low-dose irradiation in a glioma animal model, previously developed by the same team, on the G152 cell line (Beauchesne et al., 2003; Pedeux et al., 2003). Briefly the model was developed as follows; G152 malignant glioma cells (2 x106) suspended in 0.1 ml of PBS were subcutaneously injected into the inter-scapular region of 4-week-old mice (female nude mice, Swiss *nu/nu*). Drinking water was supplemented with estrone (0.1 ml/100 ml of water) until death of the animal. Two perpendicular diameters (D1 and D2) of the tumors were measured once a week and tumor volume calculated from the following equation: (D1 + D2/2)3 X (/6) (Beauchesne et al., 2003; Pedeux et al., 2003). G152 xenograft tumors were grown for 17 days, and the mice were then exposed to either 0.8 Gy per fraction (3 times per day, spaced at 4 hour intervals, 4 days per week, for 2 consecutive weeks) or to a single dose of 2 Gy (once per day, 4 days per week, for 2 consecutive weeks). Another arm of tumorbearing mice were not treated. The ultrafractionated irradiation was delivered by a clinical linear accelerator, with the mice immobilized in plastic tubes, and only the tumor exposed to

Tumors grew faster in the untreated mice with an average tumor volume at week 12 of 1223 mm3. As expected, radiation therapy had a therapeutic effect on the tumor growth resulting in an inhibition of tumor growth of 80-90 % (Beauchesne et al., 2003; Pedeux et al., 2003). At week 12, tumor volume of the mice in the ultrafractionated arm (repeated low-doses) was half that of the mice in the standard treatment arm (single dose, each day) representing a highly significant difference (p=0.0022) (Beauchesne et al., 2003; Pedeux et al., 2003). A second experiment gave similar results with neuropathology analysis revealed that the grafted tumor had the same characteristics as the initial human primary glioma tumor from

To further assert the therapeutic efficiency of ultrafractionated regimen, a third experiment was performed to compare the irradiation regimens for the same total doses (Beauchesne et al., 2003; Pedeux et al., 2003). Seventeen days after grafting, the mice were exposed to either 0.8 Gy, 3 times/day spaced at 4 hour intervals 5 days/week for 2 consecutive weeks (total dose = 24 Gy) or to 2.4 Gy once/day 5 days/week for 2 consecutive weeks (total dose = 24 Gy). Another group of mice was left untreated. Tumor size was measured once a week. As previously demonstrated, the ultrafractionated regimen led to a dramatic inhibition of tumor growth, and in the group of mice irradiated with fractions of 2.4 Gy, the tumor growth was not very different from mice irradiated with 2 Gy per fractions (Fig. 4)

effective than the conventional approach (Beck-Bornholdt et al., 1989).

the irradiation (Beauchesne et al., 2003; Pedeux et al., 2003).

which the G152 was obtained.

(Short et al., 1994a, 2001). Cell survival was calculated by DMIPS cell assay after 15 fractions of 0.4 Gy, given three times a day and compared to the same total dose given as once-daily 1.2 Gy. The low-doses were administered at 4-hours intervals (09.00 – 13.00 – 17.00 hours) each day for 5 consecutive days, and the single dose of 1.2 Gy was also given for 5 consecutive days (Short et al., 1994a, 2001). The repeated low-doses produced a significantly increased tumor cell kill; cell survival after three consecutive 0.4 Gy fractions was lower than after the same total dose given as a single fraction (1.2 Gy), the difference was significant (p<0.0002) (Short et al., 1994a, 2001). Cell survival after 2 Gy single doses was not different to that obtained after three consecutive 0.4 Gy fractions (Short et al., 1994a, 2001).Two other human glioblastoma cell lines, A7 – U87, were also tested: the lowest cell survival occurred with doses administered at 4 and 6 hours intervals for A7 and at 1 and 5 hours intervals for U87 (Short et al., 1994a, 2001). The cell line U373 (obtained from a human astrocytoma grade III) did not express HRS phenomenon, repeated low-doses did not enhance cell killing (Short et al., 1994a, 2001). The conclusions of this work were that multiple low-doses (< 1 Gy) per day spaced at appropriate intervals (4 hours) could increase cell killing by the enhancing he HRS phenomenon (Short et al., 1994a, 2001). The authors termed this multiple low-doses per fraction per day as an "ultrafractionated regimen" (Short et al., 1994a, 2001).

Fig. 3. Survival of human glioma cells following repeated irradiations. G5, CL35, G152 or MRC5 cells were exposed to 0.8 Gy 3 times/day spaced by 4 hr for 2 consecutive days or to 2 Gy once/day for 2 consecutive days. Cell survival was assessed by a clonogenic assay.

Beauchesne et al also tested the cumulative effect of low radiation doses on cell survival on the following human glioblastoma cell lines; G5 – CL35 – G152 (Beauchesne et al., 2003; Pedeux et al., 2003). Three fractions of 0.8 Gy spaced at 4 hour intervals were compared to a biologically equivalent single dose of 2 Gy. Irradiations were given for 2 consecutive days. A marked increase in cell killing was reported with the ultrafractionated regimen (repeated low-doses) in the G5 and G152 cell lines, but not in the CL35 cell line (Fig. 3) (Beauchesne et al., 2003; Pedeux et al., 2003). The experiments were repeated with a linear accelerator used daily for clinical therapies for patients. G5 and CL35 cell lines were exposed to 0.8 Gy, three times per day, spaced at 4 hour intervals for 2 consecutive days and to 2.4 Gy once a day for 2 days. Again, a marked and significant increase in cell killing occurred after the repeated low-doses for G5 but not the CL35 cell line (Beauchesne et al., 2003; Pedeux et al., 2003). It was postulated that te HRS phenomenon was responsible for the lower cell survival obtained after ultrafractionated regimen (Beauchesne et al., 2003; Pedeux et al., 2003).
