To evaluate the safety and efficacy of carbon-ion radiation therapy (C-ion RT) for locally advanced sinonasal malignant tumors in a multicenter retrospective study (J-CROS 1402 HN). Clinical data were collected for patients who had sinonasal malignant tumors of stage N0-1M0 and received C-ion RT at 4 institutions in Japan between November 2003 and December 2014. Of the 458 patients, 393 had naïve tumors and 65 had recurrent tumors. The tumors were located in the nasal cavity (n = 263), maxillary sinus (n = 109), ethmoid sinus (n = 71), and other locations (n = 15). The histologic types were mucosal melanoma (n = 221, 48%), adenoid cystic carcinoma (n = 122, 27%), squamous cell carcinoma (n = 31, 7%), olfactory neuroblastoma (n = 30, 7%), adenocarcinoma (n = 21, 5%), and other types (n = 33, 7%). Of the 458 patients, 300 (66%) had T4 tumors. All patients received definitive C-ion RT. The median follow-up period was 25.2 months for all patients (range, 1.4-132.3 months). The 2-year overall survival and local control rates were 79.6% and 84.1%, respectively. As analyzed according to histology, the 2-year overall survival rate was 68.0% for mucosal melanoma, 96.8% for adenoid cystic carcinoma, 70.0% for squamous cell carcinoma, 96.7% for olfactory neuroblastoma, and 89.7% for adenocarcinoma. Regarding late toxicities, 17% of patients developed grade 3 and 4 toxicities, of which visual impairment was the most common. The results of our multicenter study have demonstrated that C-ion RT can provide excellent clinical outcomes with acceptable late toxicities in patients who have locally advanced sinonasal malignant tumors.

The purpose of this study was to compare carbon ion radiotherapy (C-ion RT) and stereotactic radiotherapy (SBRT) with photon beams for the treatment of hepatocellular carcinoma (HCC), specifically with regard to the dose volume parameters for target coverage and normal tissue sparing. Data of 10 patients who were treated using C-ion RT with a total dose of 60 Gy(RBE) in four fractions were used. The virtual plan of SBRT was simulated on the treatment planning computed tomography images of C-ion RT. Dose volume parameters such as minimum dose covering 90 % of the planning target volume (PTV D90), homogeneity index (HI), conformity index (CI), mean liver dose (MLD), volume of the liver receiving 5 to 60 Gy (V5-60), and max point dose (Dmax) of gastrointestinal (GI) tract were calculated from both treatment plans. The PTV D90 was 59.6 ± 0.2 Gy(RBE) in C-ion RT, as compared to 56.6 ± 0.3 Gy in SBRT (p < 0.05). HI and CI were 1.19 ± 0.03 and 0.79 ± 0.06, respectively in C-ion RT, as compared to 1.21 ± 0.01 and 0.37 ± 0.02, respectively in SBRT. Only CI showed a significant difference between two modalities. Mean liver dose was 8.1 ± 1.4 Gy(RBE) in C-ion RT, as compared to 16.1 ± 2.5 Gy in SBRT (p < 0.05). V5 to V50 of liver were higher in SBRT than C-ion RT and significant differences were observed for V5, V10 and V20. Dmax of the GI tract was higher in SBRT than C-ion RT, but did not show a significantly difference. C-ion RT provides an advantage in both target conformity and normal liver sparing compared with SBRT.

This study aimed to evaluate the efficacy of carbon‐ion radiotherapy in combination with chemotherapy using dacarbazine, nimustine, and vincristine (DAV therapy) in mucosal melanoma. Twenty‐one patients with clinically localized mucosal melanoma of the head and neck were enrolled. The primary endpoint was 3‐year overall survival (OS). Secondary endpoints included local control, progression‐free survival (PFS), and adverse event occurrence. Carbon‐ion radiotherapy with a dose of 57.6‐64.0 Gy (relative biological effectiveness) in 16 fractions was delivered concurrently with DAV therapy, and 2 cycles of adjuvant DAV therapy were administered every 6 weeks. The median follow‐up periods were 15.5 months for all patients, and 31.2 months for 12 surviving patients. All patients had locally advanced T4a or T4b disease in the rhino‐sinus area. In 16 patients (76.2%), 3 cycles of planned DAV therapy were completed. The 3‐year OS and PFS rates were 49.2% and 37.0% respectively. The 3‐year local control rate was 92.3%. Eleven patients (52%) developed distant metastasis, which was the most frequent pattern of the first failure. Commonly presenting acute grade 2‐3 toxicities associated with radiotherapy and chemotherapy were mucositis (11 patients [53%]) and leukopenia (9 patients [43%]), which improved with conservative therapy. None of the patients developed grade 3 or greater late toxicities. Carbon‐ion radiotherapy in combination with DAV therapy led to excellent local control for advanced mucosal melanoma within acceptable toxicities. The efficacy of additional DAV therapy in improving survival was weaker than expected as distant metastases still occurred frequently. Trial registration no. UMIN000007939.

A single-institutional prospective study of optimal hypofractionated conformal radiotherapy for large brain metastases with high risk factors was performed based on the risk prediction of radiation-related complications. Eighty-eight patients with large brain metastases ≥10 cm3 in critical areas treated from January 2010 to February 2014 using the CyberKnife were evaluated. The optimal dose and number of fractions were determined based on the surrounding brain volume circumscribed with a single dose equivalent (SDE) of 14 Gy (V14) to be less than 7 cm3 for individual lesions. Univariate and multivariate analyses were conducted. As a result of optimal treatment, 92 tumors ranging from 10 to 74.6 cm3 (median, 16.2 cm3) in volume were treated with a median prescribed isodose of 57% and a median fraction number of five. In order to compare the results according to the tumor volume, the tumors were divided into the following three groups: 1) 10–19.9 cm3, 2) 20–29.9 cm3 and 3) ≥30 cm3. The lesions were treated with a median prescribed isodose of 57%, 56% and 55%, respectively, and the median fraction number was five in all three groups. However, all tumors ≥20 cm3 were treated with ≥ five fractions. The median SDE of the maximum dose in the three groups was 47.2 Gy, 48.5 Gy and 46.5 Gy, respectively. Local tumor control was obtained in 90.2% of the patients, and the median survival was nine months, with a median follow-up period of seven months (range, 3-41 months). There were no significant differences in the survival rates among the three groups. Six tumors exhibited marginal recurrence 7-36 months after treatment. Ten patients developed symptomatic brain edema or recurrence of pre-existing edema, seven of whom required osmo-steroid therapy. No patients developed radiation necrosis requiring surgical resection. Our findings demonstrate that the administration of optimal hypofractionated conformal radiotherapy based on the dose-volume prediction of complications (risk line for hypofractionation), as well as Kjellberg’s necrosis risk line used in single-session radiosurgery, is effective and safe for large brain metastases or other lesions in critical areas.

Lung cancer is a leading cause of cancer-related deaths worldwide. Radiotherapy is an essential treatment modality for inoperable non-small cell lung cancer (NSCLC). Stereotactic body radiotherapy (SBRT) is the standard treatment for early-stage NSCLC because of its favorable local control (LC) compared to conventional radiotherapy. Carbon ion radiotherapy (CIRT) is a kind of external beam radiotherapy characterized by a steeper dose distribution and higher biological effectiveness. Several prospective studies have shown favorable outcomes. However, there is no direct comparison study between CIRT and SBRT to determine their benefits in the management of early-stage NSCLC. Thus, we conducted a retrospective, single-institutional, and contemporaneous comparison study, including propensity score-adjusted analyses, to clarify the differences in oncologic outcomes. The 3-year overall survival (OS) was 80.1% in CIRT and 71.6% in SBRT (p = 0.0077). The 3-year LC was 87.7% in the CIRT group and 79.1% in the SBRT group (p = 0.037). Multivariable analyses showed favorable OS and LC in the CIRT group (hazard risk [HR] = 0.41, p = 0.047; HR = 0.30, p = 0.040, respectively). Log-rank tests after propensity score matching and Cox regression analyses using propensity score confirmed these results. These data provided a positive efficacy profile of CIRT for early-stage NSCLC.