Combined treatment with lexatumumab and irradiation leads to strongly increased long term tumour control under normoxic and hypoxic conditions

Purpose The combination of ionizing radiation with the pro-apoptotic TRAIL receptor antibody lexatumumab has been shown to exert considerable synergistic apoptotic effects in vitro and in short term growth delay assays. To clarify the relevance of these effects on local tumour control long-term experiments using a colorectal xenograft model were conducted. Materials and methods Colo205-xenograft bearing NMRI (nu/nu) nude mice were treated with fractionated irradiation (5× 3 Gy, d1-5) and lexatumumab (0.75 mg/kg, d1, 4 and 8). The tumour bearing hind limbs were irradiated with graded single top up doses at d8 under normoxic (ambient) and acute hypoxic (clamped) conditions. Experimental animals were observed for 270 days. Growth delay and local tumour control were end points of the study. Statistical analysis of the experiments included evaluation of tumour regrowth and local tumour control. Results Combined treatment with irradiation and lexatumumab led to a pronounced tumour regrowth-delay when compared to irradiation alone. The here presented long-term experiments revealed a highly significant rise of local tumour control for normoxic (ambient) (p = 0. 000006) and hypoxic treatment (p = 0. 000030). Conclusion Our data show that a combination of the pro-apoptotic antibody lexatumumab with irradiation reduces tumour regrowth and leads to a highly increased local tumour control in a nude mouse model. This substantial effect was observed under ambient and more pronounced under hypoxic conditions.


Background
Lexatumumab is a fully human agonistic antibody with a distinct tumour cell specifity via activation of TRAIL (TNF-related apoptosis inducing ligand) receptor 2 (TRAIL-R2) induced apoptosis. Although TRAIL-R2 stimulation alone is highly effective in a wide range of cancer cell lines, effi-cacy can be increased by combination with other gyrostatic drugs (for review see [1]). We have already shown that a combined treatment with TRAIL and irradiation exerts highly synergistic effects regarding apoptosis induction. This enhanced efficacy was detectable in various solid tumour cell lines and lymphoid tumour cells [2,3].
Combining mapatumumab or lexatumumab with irradiation, we have demonstrated that this combination exerts strong additive and synergistic effects on apoptosis induction in vitro and in short-term growth delay experiments [10]. However, to proof that induction of apoptosis evidently translates into definitive tumour stem cell eradication long-term experiments with local tumour control as primary endpoint might provide a reliable model for clinical potency [22][23][24][25][26].
Therefore, we decided to perform long-term experiments in a nude mouse xenograft model. As radiation sensitivity becomes affected by limiting intratumoural hypoxia we run experiments under both ambient and hypoxic conditions to mimic realistic tumour conditions [27].
Taken together, our experimental series was designed to confirm the striking principle that radiation mediated TRAIL sensitization effectively increases long-term local tumour control.

Animals and tumours
Immunodeficient NMRI-(nu/nu)-nude mice were purchased from a specific pathogen free colony at the University of Essen (Germany) at the age of 4-6 weeks. Animals were kept in an individually ventilated cage rack system (Techniplast, Italy) and fed with sterile high calorie laboratory food (Sniff, Germany). Drank water was supplemented by chlorotetracycline and potassium sorbate acidified to a pH of 3.0 with hydrochloric acid.
The Colo205 tumour cell line (established from a colorectal adenocarcinoma) was acquired from ATCC (Bethesda, MD, USA). In NMRI-(nu/nu)-nude mice Colo205 cells form solid, roundly shaped tumours without indication for metastasis.

Transplantation and experimental design
Tumour lumps of about 2 mm diameter from a source tumour were implanted subcutaneously into the right hind limb of 6-10 week old animals. Approximately 2-3 weeks after transplantation tumour growth was measurable. Tumour size was quantified with calipers in two perpendicular diameters. The tumour volume (V) was calculated as V = (a × b 2 )/2, where a and b are the long axis and the short axis, respectively. Scoring of tumour sizes took place three times per week before start of treatment. Body weight was monitored once a week.
The median tumour volume at the start of experiments was 116 ± 31 mm 3 . Animals were randomly allocated to 24 treatment arms (scheme see Figure 1): lexatumumab at day 1, 4 and 8 (0.75 mg/kg body weight intraperitoneally (i.p.)) alone, fractioned radiotherapy (  To minimize toxic side effects and to apply high irradiation doses in an easy comparable, time saving schedule we choose a combination of fractionated and graded single high dose (top up) irradiation. 3 Gy single dose was chosen for fractionated irradiation based on previous experiments (Marini et al., Oncogene 2006). Fractionated irradiation of tumours was applied in inhalation (Isoflurane) narcosis. Top up irradiation under ambient conditions or under clamped hypoxia was performed with i.p. narcosis (fentanyl, midazolam, medetomidine), as recommended by the university veterinarian department. For animals, whose tumours were clamped irradiation was performed 10 minutes after applying a narrow lace to the right hind limb just at the proximal end of the tumour to make the hypoxic radiation conditions as consistent as possible. Experiments were performed in one run with 252 animals.
Tumour volumes were scored twice a week, no blinding took place. Follow up was discontinued after 270 days or in case of intercurrent death or if tumours had grown to eight-times the initial tumour volume at the start of treatment. Growth delay and local tumour control were endpoints of the study. All animal experiments were accomplished in accordance with the guidelines of the local authorities (Regional Board Tuebingen, Germany, appl.no. R4/04) and the German animal welfare regulations.

Statistical Analysis
Statistical analysis was performed as described before [21]. In short terms, an exponential regression model was used to interpolate median tumour regrowth times. Regrowth delay was compared by unparametric Kruskal-Wallis tests with Dunn's post tests. Tumour control rates were calculated accounting for censored animals as described by Walker and Suit [28]. Data were analysed by a probit non linear regression analysis. Parameters were estimated using the maximum likelihood method. Statistical significance was calculated asymptotically by means of a Hessian matrix (STATISTICA 6.0 StatSoft, Hamburg, Germany).

Results
Treatment with lexatumumab failed to induce any immune reactions of the irradiated skin. No evidence of acute toxicity was observed. Follow up revealed no significant differences in frequency of intercurrent deaths after irradiation alone or combined treatment with lexatumumab (5.6% vs. 4.6%). Figure 2 shows a chronological sequence of the impressive tumour regression after treatment with lexatumumab (0.75 mg/kg) for one test animal, exemplarily. Obviously, tumour growth reduction started after the second application i.p., already. However, lacking consolidating irradiation in this example tumour regrowth is evident four weeks after start of treatment.
However, combination of very low doses of irradiation with lexatumumab led to an unexpected high local tumour rate, already. Tumour regrowth after combined treatment was observed in less than 50% of the animals. Figure 3 shows data on the 2-, 4-and 8-fold tumour regrowth after single and combined treatment with a 10 Gy top up dose, exemplarily. In this subset of experiments, five of nine mice were lacking any tumour regrowth 270 days after start of treatment. Analysis of the median time of tumour regrowth after combined treatment was impaired by an unexpected high rate of local control ( figure 3). Therefore, we decided to choose the more complex probit non linear regression analysis.  (table 1).

Discussion
Our data prove that the combination of the proapoptotic human antibody lexatumumab with ionizing radiation has an obvious influence on local tumour control in a long-term xenograft model. The effect is evident after irradiation with low doses, already.
It is important to note that these experiments with an agonistic antibody against TRAIL receptor DR5 corroborate our recently published data on a high efficacy of a combined treatment with another proapoptotic antibody (mapatumumab, anti-DR4) and irradiation. Both models are in line with in vitro data from our and other labs demonstrating that irradiation acts as a TRAIL sensitizer and not obversely [3,29,30]. This principle diverges from other combined approaches where classical chemotherapeutic or other molecular targeted agents act as radiosensitizer. E.g. the synergizing efficacy of cisplatin is based on increased oxygenation of hypoxic cells and an influence in DNA-repair and cell cycle regulation [31][32][33]. Cetuximab, an antibody against epidermal growth factor receptor, seems also to influence long-term tumour control by affecting DNA damage repair [34,35].
In contrast to former reports the mitochondrial pathway has a strong impact in TRAIL induced apoptosis. Depending on the cell system applied mitochondrial amplification loops account for its high efficacy [36,37]. In combination with TRAIL, irradiation increases apoptosis in tumour cells with an impaired mitochondrial pathway. Furthermore, preirradiation of bcl-2 overexpressing lymphoma cells raises cell death rates after TRAIL receptor stimulation [38]. In several tumour cell systems, the proapoptotic molecule Bax was shown to be essential for the combined effect of TRAIL and ionizing radiation suggest-ing a considerable mitochondrial relevance for this synergizing principle [10,39,40].
The role of radiation induced TRAIL receptor upregulation has been discussed extensively. However, we and others found an only weak or lacking correlation between upregulation and synergism [10,41,42]. Although, other mechanisms like cell cycle regulation might play a role [43].
It is important to note, that this synergistic principle works under ambient and hypoxic conditions as well. Weinmann et al. demonstrated an undiminished efficacy of TRAIL alone under hypoxia in a lymphoma cell model [44]. Takahashi at al. reported similar observations on clonogenic cell kill of A549 cells after treatment with TRAIL and irradiation [45]. However, it remains speculative why this effect on local tumour control is more pronounced under normoxia than under hypoxia. The known increase of intrinsic radioresistance of hypoxic cells will be responsible for this reduced susceptibility.
Median tumour regrowth times, calculated for two-, four-, and eight-fold tumour size of the initial tumour volume at start of treatment The strong request on the development of personalized targeted therapies has amazingly changed the general approach to cancer treatment. In contrast to cytostatic drugs being prescribed on base of classical features as TNM classification and histology, targeted drugs require an accurate identification of patient collectives who benefit from a given treatment. Therefore, a specific subset of marker molecules should be identified for each targeted drug [46][47][48].

Conclusion
The here presented data provide evidence that the combination of apoptosis inducing antibodies with irradiation strongly increases long-term tumour control. Since murine long-term control experiments are the only currently accepted functional approach to simulate the efficacy of radiation based treatments the given data are an optimal scientific base for subsequent clinical trials.

Competing interests
The authors declare that they have no competing interests.

Authors' contributions
PM conceived and drafted the manuscript. DJ and SS carried out the animal experiments to the same portion. WB performed the statistical analysis. MN participated in the statistical analysis and in the drafting of the manuscript. CB contributed to interpretation of the data and critically reviewed the article. All authors read and approved the final manuscript.