Biological effects of hydrogen peroxide administered intratumorally with or without irradiation in murine tumors

Abstract

Despite insufficient laboratory data, radiotherapy after intratumoral injection of hydrogen peroxide (H₂O₂) is increasingly being used clinically for radioresistant tumors. Especially, this treatment might become an alternative definitive treatment for early and advanced breast cancer in patients who refuse any type of surgery. The purpose of this study was to investigate the biological effects and appropriate combination methods of irradiation and H₂O₂ in vivo. SCCVII tumor cells transplanted into the legs of C3H/HeN mice were used. Chronological changes of intratumoral distribution of oxygen bubbles after injection of H₂O₂ were investigated using computed tomography. The effects of H₂O₂ alone and in combination with single or five‐fraction irradiation were investigated using a growth delay assay. The optimal timing of H₂O₂ injection was investigated. Immunostaining of tumors was performed using the hypoxia marker pimonidazole. Oxygen bubbles decreased gradually and almost disappeared after 24 h. Administration of H₂O₂ produced 2–3 days’ tumor growth delay. Tumor regrowth was slowed further when H₂O₂ was injected before irradiation. The group irradiated immediately after H₂O₂ injection showed the longest tumor growth delay. Dose‐modifying factors were 1.7–2.0 when combined with single irradiation and 1.3–1.5 with fractionated irradiation. Pimonidazole staining was weaker in tumors injected with H₂O₂. H₂O₂ injection alone had modest antitumor effects. Greater tumor growth delays were demonstrated by combining irradiation and H₂O₂ injection. The results of the present study could serve as a basis for evaluating results of various clinical studies on this treatment.

The decrease in therapeutic effects of radiotherapy for large tumors, compared with small tumors, is well known from various experimental and clinical data. One of the reasons for this decrease is the presence of hypoxic cells.1, 2 Other causes of radioresistance include the presence of anti‐oxidative enzymes such as peroxide and catalase that neutralize reactive oxygen species produced by irradiation.1 So far, various strategies to enhance radiation effects have been developed and tested in clinical trials. These include hyperbaric oxygen, nitroazole sensitizers and hypoxic cytotoxin.3, 4 Hypoxic cell radiosensitizers have been investigated extensively, but the results remain inconclusive; they may have effects when combined with high‐dose‐per‐fraction radiotherapy such as intraoperative radiotherapy,5, 6 whereas with conventional fractionated radiotherapy, it may be rather difficult to demonstrate clinical efficacy.

To overcome the radioresistance of large tumors undergoing radiotherapy, Kochi Oxydol‐Radiation Therapy for Unresectable Carcinomas (KORTUC) using hydrogen peroxide (H₂O₂) as a radiosensitizer was developed at the Department of Radiology, Kochi University, Japan.7, 8 In this treatment, H₂O₂ mixed with sodium hyaluronate is injected intratumorally two or three times every week during radiotherapy. The treatment has spread steadily and remarkable efficacy has been reported clinically.9, 10 KORTUC has shown definite effects even when combined with conventionally fractionated radiation therapy. Based on these encouraging results, a clinical trial was started also in the United Kingdom.

Despite the progress of clinical studies, however, biological data to support the efficacy of this approach is still insufficient. The investigators at Kochi University only reported partial laboratory data showing the efficacy of once‐weekly administration of H₂O₂ with gelatin‐based hydrogel (instead of sodium hyaluronate) and irradiation in tumor‐bearing mice using a small number of mice (n = 6 per group).11 However, the optimal timing of H₂O₂ administration, the magnitude of enhancement and the efficacy when combined with fractionated irradiation have not been investigated. The purpose of this study was to thoroughly investigate the biological effects and appropriate combination methods of irradiation and H₂O₂ in vivo.

Materials and Methods

Tumors, mice, hydrogen peroxide preparation and irradiation

All experiments were approved by the Animal Ethics Committee and were conducted in accordance with the principles of Nagoya City University in Japan. SCCVII cells (a squamous cell carcinoma line) and female C3H/HeN mice were used throughout the study. Characteristics of the tumors were described in detail previously.12 The treatment for the right hind legs of the mice could be most readily and quickly performed compared with that for other sites. Therefore, it was considered suitable to transplant SCCVII cells into the right hind legs of the mice. SCCVII cells cultured in vitro were transplanted subcutaneously into the right hind legs (5 × 10⁵ cells per leg) of 8‐week‐old mice. Experiments were carried out when the mean diameter of the tumors reached approximately 10 mm at 11 or 12 days after tumor cell inoculation, except for the experiments to investigate oxygen bubble distribution using computed tomography (CT). Twelve mice were used for each irradiated group. In all experiments, 0.5% w/v (0.147 mol/L) H₂O₂ (Oxydol; Ken‐ei Pharmaceutical Co. Ltd., Osaka, Japan) was prepared with sodium hyaluronate (ARTZ Dispo; Seikagaku Corporation, Tokyo, Japan) in accordance with the regimen of previous investigations.13 Sodium hyaluronate was used to relieve pain at the site of injection and preserve high intratumoral oxygen concentration.13 The tumor‐bearing legs of the mice were fixed with adhesive tape without anesthesia at the time of injection and irradiation, in accordance with the method described in detail previously;14 this method appeared not to excessively stress the mice. A thin needle (26 G needle for Tuberculin; Terumo Corporation, Tokyo, Japan) was used for injection. H₂O₂ was injected into the centers of the tumors slowly over approximately 15 s, paying attention not to cause pain to the mice. The whole body was shielded using thick lead except for the tumor‐bearing leg. Irradiation was performed using a 210‐kVp X‐ray machine (10 mA with a 2‐mm Al filter; Chubu Medical Co., Matsusaka, Japan) at a dose rate of 2.2 Gy/min as described in detail previously.15

Distribution of oxygen bubbles in tumor

First, changes of intratumoral distribution of oxygen bubbles over time after injection of H₂O₂ were investigated using a 16‐row multislice CT (Optima CT 580W; General Electric, Fairfield, CT, USA) with three mice per group. The tube voltage, tube current, field of view, and matrix size were 120 kV, 344 mA, 50.0 cm, and 512 pixels, respectively. This experiment was carried out when the mean diameter of the tumors reached about 14 mm, considering the ease of observing oxygen bubbles on CT. Three volumes (0.25, 0.5 and 1.0 mL) of 0.5% w/v H₂O₂ prepared in sodium hyaluronate were investigated. For control groups, 0.5 mL sodium hyaluronate was injected. The tumors were serially scanned until 24 h after H₂O₂ injection. The proportion of oxygen bubbles in the tumor was analyzed quantitatively on CT slices of maximal tumor size using ImageJ Version 1.49, an open source image processing software developed at the National Institutes of Health (Bethesda, MD, USA).16

Tumor growth delay assay

First, to investigate the effects of sodium hyaluronate and H₂O₂ injection alone, 0.5 mL sodium hyaluronate with or without 0.5% w/v H₂O₂ was administered intratumorally. The solutions were injected once, three times every other day, or five times every other day. To investigate the effect of sodium hyaluronate, 0.5 mL saline was injected for comparison.

Second, the combined effects of single irradiation and H₂O₂ were examined, with 0.5 mL 0.5% w/v H₂O₂ administered intratumorally. The tumors were irradiated with 18 Gy immediately (about 1 min) after injection or 15, 30, 60 or 120 min later. As a control, the tumors were also irradiated with 18 Gy after injection of 0.5 mL sodium hyaluronate.

Third, to estimate the dose‐modifying factor of this treatment, the tumors were irradiated with graded doses of 7, 14 and 21 Gy immediately after injection of sodium hyaluronate with or without H₂O₂.

Fourth, the combined effects of H₂O₂ and fractionated irradiation were examined; 2, 3, 4 or 5 Gy was administered five times over 5 days, once a day, with or without H₂O₂. The irradiation interval was 24 h. H₂O₂ was administered just before irradiation every other day (first, third and fifth days), simulating the clinical situation, for a total of three times.

In all experiments, the three dimensions of each tumor were measured every other day using a caliper. The tumor volumes were calculated as V = π/6 × products of the three dimensions. The tumor growth time (TGT) was defined as the time required for a tumor to reach 2.5 times the initial volume. The tumor growth delay time (TGDT) was calculated as the TGT in each treated group minus the TGT in the control group.

Immunofluorescence staining

To evaluate hypoxic regions in tumors, a hypoxia marker, pimonidazole, was used.17 Initially, 0.5 mL 0.5% w/v H₂O₂ was injected into 1‐cm sized tumors. After 15 min, pimonidazole HCl (Hypoxyprobe‐1; Hypoxyprobe Inc., Burlington, MA, USA) was administered intravenously to treated mice as a single 60 mg/kg dose. After 60 min, the mice were killed and the tumors were quickly removed. The tumors were fixed in 4% v/v formaldehyde, embedded in paraffin and sectioned at 5‐μm thickness with a microtome. Each slide was stained with fluorescein isothiocyanate‐conjugated anti‐pimonidazole mouse monoclonal IgG1 antibody (FITC‐MAb1 included in the Hypoxyprobe‐1 kit). Images were captured using a fluorescence microscope. The proportions of pimonidazole‐positive areas in the tumor specimen were analyzed quantitatively on the slide of maximal tumor size using ImageJ.

Statistical analysis

All statistical analyses were carried out using an open source software R Version 3.2.3 (The R Foundation for Statistical Computing, Vienna, Austria). Differences between pairs of growth delay curves were analyzed by two‐way analysis of variance, followed by post hoc Tukey's HSD (honestly significant difference) test. Differences in the proportion of oxygen bubbles on CT images between the groups injected with H₂O₂ at different timings were analyzed by Tukey's HSD test. Differences in the proportion of pimonidazole‐positive areas between the H₂O₂‐injected and control groups were analyzed by Student's t‐test.

Results

Distribution of oxygen bubbles in the tumor

Figure 1 shows CT images of the tumors at various times after injection of sodium hyaluronate with or without H₂O₂. For each group, one mouse showing representative images was chosen among the three mice per group. After injection of sodium hyaluronate alone, no bubbles were seen at 0 and 24 h after injection. On the other hand, oxygen bubbles were seen in all groups until 12 h after H₂O₂ + sodium hyaluronate injection, and at 24 h in one mouse in each group (three of the nine mice in total). In Figure 1, the proportions of bubbles in the tumors on the CT images are also shown. The proportions were largest immediately after H₂O₂ injection and decreased gradually until 24 h. Even by increasing the volume of injection, the proportion of bubbles did not necessarily increase, so 0.5 mL of H₂O₂ with sodium hyaluronate was used throughout the subsequent experiments.

Figure 1.

Changes in the intratumoral distribution of oxygen bubbles over time after injection of sodium hyaluronate with or without H₂O₂. The mean values below the images represent the percentage of oxygen bubbles on CT slices of the maximum tumor size (three mice per group). Compared with the groups receiving H₂O₂ immediately (1 min) before scanning, all the other groups had significantly smaller proportions of oxygen bubbles (P < 0.05). In the mice receiving 0.5‐mL injection, there were no differences in the proportion of oxygen bubbles among the 15‐min, 30‐min, and 60‐min groups (15 min vs 30 min: P = 0.18; 15 min vs 60 min: P = 0.057; 30 min vs 60 min: P = 0.65).

Tumor growth delay assay

Figure 2(a,b) shows tumor growth delay curves after intratumoral injection (1, 3 or 5 times) of sodium hyaluronate or H₂O₂ + sodium hyaluronate. Injection of sodium hyaluronate alone had no effect, but H₂O₂ injection produced modest growth delay, irrespective of the number of injections. Table 1 shows TGT after each treatment. The mean TGT was elongated by 2–3 days by H₂O₂ injection.

Figure 2.

Growth delay curves of SCCVII tumors. (a) Effects of sodium hyaluronate. ●, control (saline injection); , single sodium hyaluronate injection on day 0; , three sodium hyaluronate injections on days 0, 2 and 4; five sodium hyaluronate injections on days 0, 2, 4, 6 and 8. Bars represents SE of 12 mice. (b) Effects of H₂O₂. ●, control (sodium hyaluronate injection); H₂O₂ with sodium hyaluronate, a single injection on day 0; , H₂O₂ with sodium hyaluronate, three injections on days 0, 2 and 4; , H₂O₂ with sodium hyaluronate, five injections on days 0, 2, 4, 6 and 8. Bars represents SE of 12 mice. (c) Influence of the interval between H₂O₂ injection and irradiation. −, control (sodium hyaluronate); , H₂O₂ with sodium hyaluronate; , sodium hyaluronate + 18 Gy; , H₂O₂ with sodium hyaluronate + 18 Gy (1 min interval); , H₂O₂ with sodium hyaluronate + 18 Gy (15 min); , H₂O₂ with sodium hyaluronate + 18 Gy (30 min); , H₂O₂ with sodium hyaluronate + 18 Gy (60 min); , H₂O₂ with sodium hyaluronate + 18 Gy (120 min). Bars represent SE of 6 (non‐irradiated groups) or 12 mice (irradiated groups).

Table 1.

Tumor growth time (TGT)

Drug and treatment	TGT (days)		P
	Mean	SE
Control (saline)	7.5	0.3	–
Sodium hyaluronate single injection	7.2	0.3	0.97a
Sodium hyaluronate three injections	8.0	0.3	0.57a
Sodium hyaluronate five injections	8.1	0.3	0.38a
Control (sodium hyaluronate)	7.4	0.3	−
H2O2 + sodium hyaluronate single injection	9.5	0.5	0.005b
H2O2 + sodium hyaluronate three injections	10.6	0.4	<0.001b
H2O2 + sodium hyaluronate five injections	10.4	0.3	<0.001b

^aAgainst control (saline).

^bAgainst control (sodium hyaluronate).

Figure 2(c) shows TGDTs after a single 18‐Gy irradiation with or without prior H₂O₂ injection at various intervals. Table 2 (first column) shows TGDT after each treatment. In the groups receiving H₂O₂ injection at 1, 15, 30, 60 or 120 min before irradiation, the tumor regrowth was significantly elongated, compared with the control 18‐Gy group. The group irradiated immediately (1 min) after the injection showed the largest tumor growth delay, but there were no differences among the 15‐, 30‐, and 60‐min groups (15 min vs 30 min: P = 0.79; 15 min vs 60 min: P = 0.69; 30 min vs 60 min: P = 0.98). In all subsequent experiments, H₂O₂ + sodium hyaluronate was injected immediately before irradiation.

Table 2.

Tumor growth delay time (TGDT) after various treatments

After H2O2 ± 18 Gy	TGDT (days)		P
	Mean	SE	a	a
H2O2 + sodium hyaluronate	3.4	0.8	−	−
Sodium hyaluronate → 18 Gy	12.2	0.7	−	−
H2O2 – 1 min → 18 Gy	21.7	0.8	<0.001	−
H2O2 − 15 min → 18 Gy	16.5	1.2	0.006	0.002
H2O2 − 30 min → 18 Gy	17.7	1.2	<0.001	0.014
H2O2 − 60 min → 18 Gy	18.0	1.3	<0.001	0.030
H2O2 − 120 min → 18 Gy	16.0	1.4	0.031	0.003

	Mean	SE	P
After H2O2 + graded single doses
H2O2 + sodium hyaluronate	2.9	0.4	−
Sodium hyaluronate + 7 Gy	6.6	0.3	−
Sodium hyaluronate + 14 Gy	9.0	0.3	−
Sodium hyaluronate + 21 Gy	14.3	0.5	−
H2O2 + 7 Gy	10.6	0.5	<0.001
H2O2 + 14 Gy	15.3	0.4	<0.001
H2O2 + 21 Gy	26.2	0.7	<0.001
After H2O2 + 5‐fraction irradiation
H2O2	2.1	0.3	−
Sodium hyaluronate + 3 Gy × 5	8.6	0.6	−
Sodium hyaluronate + 4 Gy × 5	14.3	0.8	−
Sodium hyaluronate + 5 Gy × 5	25.6	0.7	−
H2O2 + 2 Gy × 5	7.6	0.6	−
H2O2 + 3 Gy × 5	15.7	0.8	<0.001
H2O2 + 4 Gy × 5	23.3	1.0	<0.001

P‐values are against the groups receiving the same doses without H₂O₂.

Against 18 Gy.

^aAgainst H₂O₂ – 1 min → 18 Gy.

Figure 3(a) shows tumor growth delay curves after single graded doses of irradiation (7, 14 and 21 Gy) with or without H₂O₂. Figure 3(b) and Table 2 (second column) show TGDT. In all groups injected with H₂O₂, the tumor regrowth was significantly elongated, compared with the groups receiving sodium hyaluronate injection + irradiation. From Figure 3(b), the overall dose‐modifying factors appeared to be 1.7–2.0 depending on the dose level. Figure 3(c) shows tumor growth delay curves after fractionated irradiation with or without prior H₂O₂ injection. Figure 3(d) and Table 2 (third column) show TGDT. In the 3 × 5 Gy and 4 × 5 Gy groups injected with H₂O₂, the tumor regrowth was significantly elongated compared with the group receiving irradiation without H₂O₂. From Figure 3(d), the dose‐modifying factors appeared to be 1.3–1.5 depending on the dose level.

Figure 3.

Growth delay curves and times of SCCVII tumors. (a) Effects of H₂O₂ injection plus graded doses of irradiation. −, control (sodium hyaluronate); , H₂O₂ with sodium hyaluronate; , sodium hyaluronate + 7 Gy; , H₂O₂ with sodium hyaluronate + 7 Gy; , sodium hyaluronate + 14 Gy; , H₂O₂ with sodium hyaluronate + 14 Gy; , sodium hyaluronate + 21 Gy; , H₂O₂ with sodium hyaluronate + 21 Gy. (b) Growth delay time as a function of radiation dose: , sodium hyaluronate + radiation; , H₂O₂ with sodium hyaluronate + radiation. (c) Effects of H₂O₂ injection plus fractionated irradiation. −, control (sodium hyaluronate); , H₂O₂ with sodium hyaluronate; , sodium hyaluronate + 3 Gy × 5 times; , H₂O₂ with sodium hyaluronate + 2 Gy × 5 times; , sodium hyaluronate + 4 Gy × 5 times; , H₂O₂ with sodium hyaluronate + 3 Gy × 5 times; , sodium hyaluronate + 5 Gy × 5 times; , H₂O₂ with sodium hyaluronate + 4 Gy × 5 times. All sodium hyaluronate with or without H₂O₂ was injected three times on days 0, 2 and 4. All irradiation was delivered five times on days 0–4. (d) Growth delay time as a function of total radiation dose. , sodium hyaluronate injection + radiation; , H₂O₂ with sodium hyaluronate injection + radiation. (a, c) Bars represent SE of 6 (non‐irradiated groups) or 12 mice (irradiated groups). (b, d) Bars represent SE. Curves were drawn by the quadratic polynomial approximate method.

Immunofluorescent staining

Figure 4 shows the immunofluorescent staining of SCCVII tumors. The centers of untreated tumors were clearly stained with pimonidazole, whereas the staining was weaker in tumors injected with H₂O₂. In the tumors injected with H₂O₂, the proportion of pimonidazole‐positive areas was significantly lower than that of the control group (n = 4/group, 3.3 ± 0.3% vs 6.1 ± 0.7%, P = 0.001).

Figure 4.

Immunofluorescent staining of SCCVII tumors at 60 min after intravenous pimonidazole administration. The blue areas represent tumors, and the green areas represent pimonidazole‐positive cells. (a) Control (sodium hyaluronate injection 15 min before pimonidazole). (b) H₂O₂ with sodium hyaluronate injection 15 min before pimonidazole administration.

Discussion

Although radiotherapy with H₂O₂ injection (KORTUC treatment) for large tumors has been used clinically, experimental data on appropriate combination methods have been lacking. This is the first detailed laboratory study to systematically investigate the in vivo efficacy of the treatment and appropriate combination methods. In applying this new treatment, the quality assurance of intratumoral H₂O₂ injection and its safety may be the issues that should be clarified. In the present study, H₂O₂ was injected from above into the center of the tumor in all tumors, so the quality of H₂O₂ injection may be assured. Clinically, the quality assurance of injection may be more difficult due to the variety of tumor site, size, and shape. To solve the problem, we are using CT‐guided injection in all cases, and other investigators use ultrasound‐guided injection. In previous publications as well as in our preliminary clinical experiences, the safety of KORTUC treatment has almost been established.(7, 8, 9, 10, 11, 18)

Intratumoral oxygen pressures at 1 or 24 h after H₂O₂‐containing sodium hyaluronate injections were reported previously.13 Oxygen pressures increased to 1400–1600 mmHg 1 h later and decreased to 80–100 mmHg at 24 h. Our CT study results seemed to be consistent with the previous results. Sufficient oxygen bubbles were detected on CT images at 0–1 h. In contrast, oxygen bubbles were almost undetected 24 h later. This result could suggest the optimal timing of injection, which is discussed later. Oxygen bubble distribution on CT has also been reported in human cancers recently.19 We used 0.5 mL of H₂O₂ + sodium hyaluronate, while clinically 3–6 mL is injected to human tumors. Since the volume of the SCCVII tumors in this study is about one‐fifth or smaller compared to human tumors, the ratio of the volume of H₂O₂ solution to the tumor volume used in this study may not greatly deviate from that used in clinics.

Various functions of H₂O₂ have been reported in literature. It was reported that DNA damage caused by H₂O₂ induces cell death by apoptosis or necrosis.20, 21 It was also reported that lysosomal rupture caused by H₂O₂ induces apoptosis.22 Modest cytotoxicity of H₂O₂ in tumor cells was confirmed in this study. Contrary to our expectations, there was no additional tumor growth delay by increasing the number of H₂O₂ injections. In other words, cell damage in the center of the tumor was induced by a single H₂O₂ injection, and additional H₂O₂ injections into the damaged regions were not effective. In the immunofluorescence staining, the centers of the tumors injected with H₂O₂ were partially destroyed, and the weaker pimonidazole staining and decrease of pimonidazole‐positive areas suggested the improvement of hypoxic conditions. In the present study, H₂O₂ injection was performed at 48 h intervals, and it is not known whether repeated H₂O₂ injection at shorter or longer intervals might induce additional cell death, and this should be investigated in the next study. If additional cell death is observed using longer intervals, repeat H₂O₂ injection over several weeks could be efficient as a cytotoxic treatment. In addition, while H₂O₂ was always injected into the center of the tumors in this study, it may be clinically possible to inject H₂O₂ to slightly different sites within the tumor each time when the tumor is large. By doing so, the effect of H₂O₂ itself may be better utilized. This should be a topic of future investigation.

Although a single H₂O₂ injection proved to have modest cytotoxic effects, combination with irradiation appeared to produce additional sensitizing effects, because the combined effect was dependent on the interval between H₂O₂ injection and irradiation. In contrast to the clinical experiences, only modest tumor shrinkage was observed in our experiments, but this is due to the radioresistance of SCCVII tumors that show marked shrinkage at doses >22.5 Gy.23 So, tumor control experiments using higher doses may be necessary in future. Nevertheless, this combined effect is probably mostly because of the increase in oxygenated cells, considering the results of pimonidazole staining. Overall dose‐modifying factors appeared to be 1.7–2.0 when combined with single high doses and 1.3–1.5 when combined with fractionated irradiation. With these levels of dose‐modifying factors, it is reasonable that definite clinical effects are demonstrated. Regarding the optimal timing of H₂O₂ injection, irradiation immediately after injection appeared to yield the best results. This was consistent with the observation on CT images. Although it may be difficult to inject H₂O₂ 1 min before irradiation, it is recommended to inject H₂O₂ as shortly before irradiation as possible. In recent clinical practice, H₂O₂ is most often delivered twice a week (on Monday and Wednesday) while radiotherapy is given 5 days a week. This is because everyday injection is labor‐intensive and uncomfortable to patients. However, based on the distribution of oxygen bubbles and biological effects, such an administration schedule may not be optimal. This should be a topic of further investigation whether everyday injection yields higher effects.

Despite the paucity of laboratory data, clinical use of H₂O₂ before radiotherapy is rapidly spreading in Japan. In addition, phase I followed by phase II clinical studies were started in 2016 at the Royal Marsden Hospital, London (Y. Ogawa, personal communication, March 2017). There have been no randomized studies, but phase II studies in operable breast cancer patients have indicated definite efficacy.18 The investigators suggested that this treatment might become an alternative definitive treatment for early and advanced breast cancer in patients who refuse any type of surgery. Efficacy against other tumors has also been reported.10 We have also started a clinical study to evaluate the safety and efficacy of treatment after approval of the institutional review board. We are mainly using this treatment for breast cancer patients who refuse surgery. Preliminary results are encouraging, and the effect appears to be apparently stronger than that obtained by radiation alone. The KORTUC treatment seems to be a promising new radiosensitization modality for locally advanced non‐deep‐seated tumors.

In conclusion, this study showed the in vivo efficacy of radiotherapy combined with prior intratumoral H₂O₂ injection. A dose‐modifying factor of 1.3–1.5 would be expected when combined with fractionated radiotherapy. The results of the present study could serve as a basis for evaluating results of various clinical studies on this treatment that are already ongoing.

Disclosure Statement

The authors have no conflict of interest.

Cancer Sci 108 (2017) 1787–1792