If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author. Address for correspondence: Naoyoshi Yamamoto, MD, PhD, National Institute of Radiological Sciences, National Institute for Quantum and Radiological Science and Technology, 4-9-1 Anagawa, Inage-ku, Chiba, 263-8555, Japan.
Our objective was to report initial results of a dose escalation trial of single-fraction carbon ion radiotherapy for peripheral stage I NSCLC.
Methods
Between April 2003 and February 2012, a total of 218 patients were treated. The total dose was raised from 28 to 50 Gy (relative biological effectiveness [RBE]). There were 157 male and 61 female patients, with a median age of 75 years. Of the tumors, 123 were stage T1 and 95 were stage T2. A total of 134 patients (61.5%) were medically inoperable. By histological type, there were 146 adenocarcinomas, 68 squamous cell carcinomas, three large cell carcinomas, and one mucoepidermoid carcinoma.
Results
The median follow-up was 57.8 months (range 1.6–160.7). The overall survival rate at 5 years was 49.4%. The local control (LC) rate was 72.7%. A statistically significant difference in LC rate (p = 0.0001, log-rank test) was seen between patients receiving 36 Gy (RBE) or more and those receiving less than 36 Gy (RBE). In 20 patients irradiated with 48 to 50 Gy (RBE), the LC rate at 5 years was 95.0%, the overall survival rate was 69.2%, and the progression-free survival rate was 60.0% (median follow-up was 58.6 months). With dose escalation, LC tended to improve. As for adverse lung and skin reactions, there were no patients with grade 3 or higher reactions, and less than 2% had a grade 2 reaction. Regarding chest wall pain, only one patient had grade 3 late toxicity.
Conclusions
We have reported the outcome of a dose escalation study of single-fraction carbon ion radiotherapy for stage I NSCLC, showing the feasibility of obtaining excellent results comparable to those with previous fractionated regimens.
Lung cancer continues to be the most common cancer in the world in terms of both the number of reported cases and mortality. Patients with NSCLC are classified into two groups for radiotherapy (RT). The first group comprises those with advanced lung cancer, including patients with invasion of the chest wall and/or mediastinal lymph nodes. The other group includes patients with early-stage disease (i.e., peripherally localized T1 or T2a tumors without evidence of lymph node metastases). In general, only patients with stage I lung cancer are expected to have a long survival. Surgical resection has played a pivotal role in the treatment of peripherally localized lung cancer, and it can provide a 5-year survival rate of 60% and a 5-year local control (LC) rate of more than 80%. As a result, the first recommended treatment for early-stage lung cancer has been surgical resection. However, this is not always feasible, and surgery may have increased morbidity owing to certain medical conditions such as pulmonary or cardiovascular disease. RT is the principal treatment option for patients with early-stage lung cancer and contraindications for surgery. The outcome from conventional radiation therapeutic techniques has been a 5-year LC rate of 40% to 70%. However, an LC rate equivalent to that with surgery is being reported thanks to recent advances in irradiation techniques, which include stereotactic body RT (SBRT), proton beam therapy, and carbon ion RT (CIRT).
Outcome in a prospective phase II trial of medically inoperable stage I non-small cell lung cancer patients treated with stereotactic body radiotherapy.
Clinical trials for various types of tumors were initiated at the National Institute of Radiological Sciences (NIRS) in June 1994. Carbon ion beams were used, and dose fractionation suitable for individual diseases and irradiation techniques such as respiratory-gated RT were developed. As a result, the healing of refractory cancers such as sarcoma of the bone and soft tissue, for which surgery is difficult, and healing of postoperative local recurrence of rectal cancer, etc., were achieved, and it was found that safe treatment over increasingly shorter periods is possible for cancers of the prostate gland, head and neck, lung, and liver.
CIRT for NSCLC was initiated in November 1994. Regarding peripheral stage I lung cancer, the treatment period and fractionation were shortened and decreased from 18 fractions over the course of 6 weeks to nine fractions over 3 weeks, and then further to four fractions over 1 week, all the while confirming safety and efficacy.
Preoperative carbon ion radiotherapy for non-small cell lung cancer with chest wall invasion—pathological findings concerning tumor response and radiation induced lung injury in the resected organs.
From April 2003, as a final step, a clinical trial consisting of irradiation being completed in a single day was conducted. The total dose was raised from 28 Gy (RBE), and finally, treatment with 50 Gy (RBE) as a single fraction could be provided and we decided to use this treatment protocol with 50 Gy (RBE) in a phase II trial.
One of the obvious advantages of completing the treatment in a single day is the convenience for the patients. Of course, another plus is the resulting cost reduction for this treatment. In addition, we expect that many patients will be able to receive this therapy comfortably. In this article, we describe the results of our dose escalation trial, especially in terms of LC and survival of patients who received higher-dose single-fraction CIRT.
Materials and Methods
Study Design
This study was a prospective clinical trial with primary end points of acute toxicity and LC rate. Secondary end points were late toxicity and survival rates. This protocol (0201) was designed by the Working Group for Lung Cancer, and the study was approved by the institutional review board of NIRS and conducted in accordance with the ethical standards of the Declaration of Helsinki.
This was a phase I/II study in which a dose escalation method was used to determine the optimal dose. The initial treatment dose was 28 Gy (RBE) administered in a single fraction by using respiratory-gated and four-portal oblique irradiation directions, with the total dose being escalated to a maximum dose of 50 Gy (RBE) at increments of 2.0 Gy (RBE). Dose-limiting toxicity (DLT) was defined as any acute grade 3 or higher toxicity. If there was no acute DLT in at least six patients for 6 months after the start of CIRT, the dose was raised to the next level.
Patients
During the 9 years from April 2003 to February 2012, a total of 218 patients were treated. Patients with a peripheral type of stage I NSCLC were eligible for this study. Peripherally located type means that the primary tumor is located at least 1 cm from the lobar bronchus. If the planning target volume included lobar bronchi even partially, we excluded the case from single-fraction irradiation. In addition, the primary tumor had to be histologically proved and measurable. Computed tomography (CT) scans of the chest and whole abdomen, enhanced magnetic resonance imaging of the brain, bone scans, and bronchial endoscopy were routinely performed to permit staging. Clinical staging was performed according to the sixth lung cancer TNM classification (Union for International Cancer Control, 2002).
Therefore, in this study, as a matter of course, T2 tumors included those that were larger than 5 cm in greatest diameter as long as they were stage I lung cancer. Further, we also reported the subclassifications of T factor as T1a, T1b, T2a, and T2b, conforming to the new (seventh) edition of the Union for International Cancer Control classification (2009).
The IASLC lung cancer staging project: proposals for the revision of the T descriptors in the forthcoming seventh edition of the TNM classification for lung cancer.
Patient characteristics are shown in Table 1. A total of 123 patients were in stage IA and 95 were in stage IB. By histological type, there were 146 adenocarcinomas, 68 squamous cell carcinomas, three large cell carcinomas, and one mucoepidermoid carcinoma. The median age was 75 years (range 46–89), and the breakdown by sex was 61 female and 157 male patients. Of these patients, 134 were medically inoperable and the other 84 refused surgery. Of the 123 patients in stage T1, 45 were in stage T1a and 78 were in stage T1b. Of the 95 patients in stage T2, 87 were in stage T2a and eight were in stage T2b.
The patients’ performance status was between 0 and 2 according to the WHO criteria. The medical history of all patients, including several factors such as age, pulmonary function, and cardiac function, was investigated by a surgeon to determine eligibility for surgery. Medically inoperable patients were those diagnosed by the surgeon as unsuitable for surgery at entry. Patients were ineligible if they had undergone previous RT to the target or had prior chemotherapy. For eligibility, patients had to be without interactive infection, interstitial pneumonitis, and active multiple cancers.
Informed consent was obtained from all patients before treatment.
RT Technique and Planning
Carbon ion beams with 290, 350, and 400 MeV of nucleon energy were generated by the Heavy-Ion Medical Accelerator in Chiba synchrotron and delivered by its transport system to the irradiation system in the treatment room. In this system, the beams are broadened and shaped to a three-dimensional tumor contour. For three-dimensional dose delivery, the key technology to cover the tumor thickness is the spread-out Bragg peak (SOBP) of the carbon ion beam by the use of a range modulator. The reference point was the center of the SOBP. CT planning was performed by using the HIPLAN system, which was specifically developed at NIRS for three-dimensional treatment planning with respiratory-gated CT images.
Patient positioning and verification were performed with patient set-up devices, digitally reconstructed radiographs, online portal fluoroscopic radiographs, and metal markers made of iridium wire as landmarks. A respiratory-gated irradiation system was developed and used to minimize respiratory movements of the tumor and to reduce treatment volume. The timing of gate on was set at the end of the expiratory phase because the motion of the diaphragm is slower and more reproducible than during the inspiratory phase.
An individual immobilization device (Moldcare [Alcare, Tokyo, Japan] or Shelfitter [Kuraray, Osaka, Japan]) was used to place patients in a supine or prone position. Planning CT images in the expiratory phase were acquired at a 5-mm slice thickness. Targets are typically irradiated from four oblique directions without prophylactic elective nodal irradiation.
A margin greater than 10 mm was set outside the gross tumor volume to determine the clinical target volume (CTV). Spicular formations and pleural indentations are included in the CTV where possible. An internal margin (IM) was set outside the CTV to allow for target motion during gating. The planning target volume was defined as the CTV plus the IM. The IM was determined by extending the target margin in the head and tail directions by a width of 5 mm.
The carbon ion radiation dose was expressed in terms of Gy (RBE). Gy (RBE) values were calculated by multiplying the physical dose by the RBE, which is approximately 3.0 at 0.8 cm from the distal end of the SOBP.
As part of follow-up, almost all of the patients underwent clinical examinations and CT scans at our institute. Clinical outcomes of all patients were confirmed. The first follow-up examinations were performed 4 weeks after CIRT. The examinations were then repeated about every 3 months for at least 2 years, after which check-ups were performed at 6-month intervals over the course of 5 years. After that, information was gathered at least once a year.
Local recurrence was defined as a lastingly enlarging tendency of the tumor at intervals of 3 months, as well as on the basis of findings of CT images, positron emission tomography scans, tumor marker levels, and biopsy results.
Toxicities were assessed according to the National Cancer Institute Common Toxicity Criteria version 2.0 (early) and the Intergroup Radiation Therapy Oncology Group criteria (late).
After therapy, the patients' progress was verified twice a year by the members of the Working Group for Lung Cancer, whose names are listed in the “Acknowledgments” section.
Statistical Analysis
Statistical analysis was performed by using the StatView (version 5.0) software package (Abacus Concepts, Berkeley, CA). LC and survival were assessed by the Kaplan-Meier method. For statistical testing, the log-rank test was used. A p value less than 0.05 was considered statistically significant.
Results
The median observation period was 57.8 months (range 1.6–160.7); the statistical 3-year LC and overall survival (OS) rates of the 218 patients were 77.9% and 68.3%, respectively; and the 5-year LC and OS rates were 72.7% and 49.4%, respectively.
Regarding follow-up, the observation period ranges by respective doses were as follows: 28 Gy (RBE), 26.3 to 160.7 months; 32 Gy (RBE), 3.0 to 152.4 months; 34 Gy (RBE), 8.0 to 144.7 months; 36 Gy (RBE), 22.1 to 134.9 months; 38 Gy (RBE), 1.6 to 120.3 months; 40 Gy (RBE), 15.2 to 124.6 months; 42 Gy (RBE), 7.8 to 120.6 months; 44 Gy (RBE), 6.7 to 116.9 months; 46 Gy (RBE), 7.3 to 91.0 months; 48 Gy (RBE), 21.9 to 69.1 months; and 50 Gy (RBE), 39.6 to 63.6 months. All information was gathered within 1 year after the previous follow-up for all patients.
For determination of the influence of radiation dose on LC, patients receiving 34 Gy (RBE) or less were compared with those receiving 36 Gy (RBE) or more. A statistically significant difference (p = 0.0001, log-rank test) was seen between the two groups (Fig. 1). The same comparisons regarding progression-free survival (PFS) and OS between the two groups also showed significant differences (Fig. 2). For further investigation of LC in terms of T factor and doses received, patients were divided into six groups according to dose: 28 to 34 Gy (RBE), 36 to 42 Gy (RBE), or 44 to 50 Gy (RBE) for each T factor (T1/T2). The resulting 3-year LC rates for 28 to 34 Gy (RBE), 36 to 42 Gy (RBE), and 44 to 50 Gy (RBE) for T1 were 80.7% (95% confidence interval [CI]: 66.8–94.6), 88.0% (95% CI: 78.2–97.8), and 90.8% (95% CI: 82.2–99.4), respectively, and for T2, they were 47.3% (95% CI: 29.7–64.9), 74.3% (95% CI: 54.7–93.9), and 77.7% (95% CI: 64.0 to 91.4), respectively. Although there were no significant differences, LC tended to improve with dose escalation (Fig. 3).
Figure 1Outcomes after single-fraction carbon ion radiotherapy (CIRT) for local control (LC) in patients receiving 36.0 to 50.0 Gy (relative biological effectiveness [RBE]) versus 28.0 to 34.0 Gy (RBE).
Figure 3Outcomes of local control (LC) separated by T1 (A) or T2 (B) and individual prescribed doses for three groups: 28.0 to 34.0, 36.0 to 42.0, and 44.0 to 50.0 Gy (relative biological effectiveness [RBE]). Abbreviation: CIRT, carbon ion radiotherapy.
In the cases receiving 44 Gy (RBE) or more, we investigated LC rate in relation to T factors: T1a of 2 cm or less, T1b of 2 to 3 cm, T2a of 3 to 5 cm, and T2b larger than 5 cm. There were no primary tumor failures in T1a. The 3-year LC rate for T1b and T2a was nearly 85%. In T2b cases, no tumor was controlled (Fig. 4).
Figure 4Local control (LC) after carbon ion radiotherapy (CIRT) for tumor size (T1a, T1b, T2a, and T2b) in patients receiving 44.0 to 50.0 Gy (relative biological effectiveness [RBE]).
As for the clinical results of the 84 patients (52 males and 32 females) who received 44 Gy (RBE) or more, median follow-up time was 62.7 months (range 6.7–116.9), 44 were T1, and 40 were T2, and the 5-year LC, OS, and PFS rates were 79.9%, 61.5%, and 43.8%, respectively. Concerning the difference in results between medically inoperable and operable patients, for instance, OS in the operable patients was superior to that in the inoperable patients in the 44–Gy (RBE) or more group, although LC was similar.
In 20 patients (T1/T2 ratio 13:7) irradiated with 48 to 50 Gy (RBE), the LC rate at 5 years was 95.0%, the OS rate was 69.2%, and the PFS rate was 60.0% (median follow-up was 58.6 months).
As for failure after CIRT, reirradiation by CIRT or salvage surgery for local recurrence or metastases was performed if possible.
Toxicities caused by CIRT were assessed according to the National Cancer Institute Common Toxicity Criteria (early) and the Intergroup Radiation Therapy Oncology Group criteria (late). All early and late reactions of lung and skin were grade 2 or less, and the rate of grade 2 reactions was less than 2% (Tables 2 and 3). Grade 3 chest wall pain, however, was found in only one patient receiving 50 Gy (RBE). Thus, except for this case, all other patients had grade 2 or lower toxicities of the chest wall, and there were seven grade 2 cases. Regarding rib fracture, no toxicities greater than grade 2 were observed.
Table 2Lung Reaction
Dose, Gy (RBE)
Early vs. Late
Case
Early (NCI-CTC)
Case
Late (RTOG/EORTC)
0
1
2
3
≤4
0
1
2
3
≤4
28.0–42.0
134
78
55
1
0
0
128
14
114
0
0
0
44
44
6
37
1
0
0
43
2
40
1
0
0
46
20
0
20
0
0
0
19
0
19
0
0
0
48
10
0
8
2
0
0
10
0
10
0
0
0
50
10
0
10
0
0
0
10
0
10
0
0
0
Total
218
84
130
4
0
0
210
16
193
1
0
0
NCI-CTC, National Cancer Institute Common Toxicity Criteria; RTOG, Radiation Therapy Oncology Group; EORTC, European Organization for Research and Treatment of Cancer; RBE, relative biological effectiveness.
NCI-CTC, National Cancer Institute Common Toxicity Criteria; RTOG, Radiation Therapy Oncology Group; EORTC, European Organization for Research and Treatment of Cancer; RBE, relative biological effectiveness.
In this report, LC of all patients receiving 36 Gy (RBE) or more showed a good result. As stated in previous reports, improvement of primary tumor control had led to enhanced survival rates in terms of PFS or OS with 36 Gy (RBE) or more.
With 36 Gy (RBE) or higher, although a significant difference between LC of the higher-dose group (44–50 Gy [RBE]) compared with LC of the lower-dose group (36–42 Gy [RBE]) could not be shown, at least LC showed a tendency to improve with dose escalation. We thus concluded that there was a promise of success, and despite the absence of DLTs, we stopped dose escalation at 50 Gy (RBE). Thus, we considered that the optimal dose, the highest dose in this study, was 50 Gy (RBE).
Regarding treatment dose, in single-fraction SBRT, it was reported that better LC was achieved in patients receiving 26 Gy or more than in those who received less than 26 Gy.
There may be a certain threshold dose above which clinical results such as LC might display impressive progress. In our CIRT, we tested every dose and decided on 36 Gy (RBE) as the threshold dose. It is difficult to show the distinction of higher doses above this threshold dose, as the difference in LC results is too small. Our results also could not provide definite evidence of an advantage for LC in the higher-dose subgroup (>36 Gy [RBE]), although we do think that there is an improvement in LC with dose escalation.
With 44 Gy (RBE) or more, LC of T1a tumor was very good. LC of T1b tumors was similar to the T2a result (about 80%). At 50 Gy (RBE), even for T2a tumor, we would expect LC to be higher than at least 80%, which would be clinically acceptable. LC of T2b was not good, but as the number of cases was small and there were no results based on treatment with 48 Gy (RBE) or more, in this article we cannot pass judgment on the feasibility of single-fraction irradiation for tumors larger than 5 cm. Increases of treatment dose may improve the result.
We are currently treating tumors 5 cm or smaller with a regimen of single-fraction CIRT with 50 Gy (RBE). In contrast, for tumors larger than 5 cm, we are applying a regimen such as that being used for locally advanced NSCLC, for which a dose escalation study using 16 fractions over 4 weeks was started in 2000, and the optimal dose was found to be 72 Gy (RBE).
Regarding operable patients, we had not initially intended this therapy for patients who could undergo surgical resection. However, the number of operable patients refusing surgery and thus receiving CIRT is on the increase. In fact, there appears to be an improving tendency in the OS rate of patients treated with radiation therapy for good performance status,
Concerning toxicity, there were no adverse lung or skin reactions higher than grade 3. We also reported the pulmonary functions and histological findings of the patients after CIRT,
Preoperative carbon ion radiotherapy for non-small cell lung cancer with chest wall invasion—pathological findings concerning tumor response and radiation induced lung injury in the resected organs.
in which normal tissue damage was held at low grade by the excellent dose distribution of carbon ion beams because of formation of the Bragg peak. It is believed that the advantages stem from adopting respiratory-gated, four-direction irradiation. There were no reactions higher than grade 3, and only one patient had grade 3 chest wall pain at 6 months after treatment with 50 Gy (RBE). The patient was a 62-year-old man with stage IB adenocarcinoma. Because he required a transdermal fentanyl patch, his reaction was assessed as grade 3. But other toxicities were grade 2 or lower. It is believed that the toxicity of single-fraction CIRT could be acceptable. Additionally, in a subgroup analysis of this dose escalation study, treatment was performed safely even for elderly patients.
A limitation of this study was that the case numbers differed in terms of treatment doses and T factors. Treatment dose was gradually increased over time during the course of the study, with the numbers of patients in the respective T-factor groups varying throughout and thus making statistical analysis difficult. Because of many eligible cases being admitted in a short period or, in contrast, few cases during an extended period, we could not equalize the patient numbers for each of the treatment doses.
It had been a matter of importance as to whether the use of a regimen of completion of irradiation in single a day would make it possible to achieve results comparable with those of fractionated regimens. Prospective trials of single-fraction SBRT for early-stage lung cancer or metastatic lung tumor have also been conducted without serious side effects.
But most reports have not shown long-term results of OS or LC. Our study has demonstrated the long-term clinical efficacy of single-fraction CIRT for stage I NSCLC. If the treatment of lung cancer can be completed in a single day, then it will be convenient and socially acceptable for patients. As it has often been indicated that CIRT is expensive compared with SBRT, one of the reasons for us to try to reduce the fraction size of CIRT for many tumors was that shortening the treatment period would be economically beneficial for every patient.
At the head of the results, we wish to emphasize the effectiveness of single-fraction CIRT by showing the clinical results of 20 patients who received 48 or 50 Gy (RBE). In spite of their short follow-up period, the preliminary results of those treated with 50 Gy (RBE) are as follows: in 40 patients (T1/T2 ratio 27:13), the 2-year LC rate was 96.7% (100% for T1 and 90% for T2:), the 2-year OS rate was 93.7%, and the median follow-up was 27.4 months (range 5.1–63.2).
As has been well known, there is size dependency in tumor control by RT. Concerning SBRT, LC for T2 tumor is lower than for T1.
A photon study of a large number of patients reported that LC for T2 is associated with a higher BED10. Further, improvement in LC was observed in patients with T2 lesions treated with a BED10 greater than 05 Gy. But LC, even if improved, was less than 80% at 3 years.
Therefore, in the present single-fraction irradiation trial, we also expected the T2 tumor control rate to exceed 80%. Furthermore, we think that we can prove the advantage of CIRT for attaining superior LC for T2 tumors compared with photon therapy.
Although more detailed analysis of the subsequent phase II results will be required, we believe that favorable results of single-fraction CIRT with the use of 50 Gy (RBE) that are comparable to the results of previous fractionated CIRT will be achieved.
Acknowledgments
This study was supported by the National Institute of Radiological Sciences under its research project on heavy ion at the National Institute of Radiological Sciences. The study contributors were the members of the Working Group for Lung Cancer, who are as follows: Norihiko Ikeda, Akira Iyoda, Yoshinori Okada, Hideki Kimura, Tetsuro Kodama, Tomoyuki Goya, Yoshiyuki Shioyama, Yuichi Takiguchi, Koichiro Tatsumi, Ryosuke Tuchiya, Yukio Nakatani, Takashi Nakano, Haruhiko Nakayama, Yuko Nakayama, Masahiko Higashiyama, Takashi Yamashita, Tetsuo Yamaguchi, and Ichiro Yoshino.
References
Baumann P.
Nyman J.
Hoyer M.
et al.
Outcome in a prospective phase II trial of medically inoperable stage I non-small cell lung cancer patients treated with stereotactic body radiotherapy.
Preoperative carbon ion radiotherapy for non-small cell lung cancer with chest wall invasion—pathological findings concerning tumor response and radiation induced lung injury in the resected organs.
The IASLC lung cancer staging project: proposals for the revision of the T descriptors in the forthcoming seventh edition of the TNM classification for lung cancer.
33581-X&id=gr1.jpg){kind=link}
33581-X&id=gr2.jpg){kind=link}
33581-X&id=gr3.jpg){kind=link}
33581-X&id=gr4.jpg){kind=link}