Clinical Investigation
Tumor Control and Toxicity for Common Stereotactic Body Radiation Therapy Dose-Fractionation Regimens in Stage I Non-Small Cell Lung Cancer

https://doi.org/10.1016/j.ijrobp.2017.10.037Get rights and content

Purpose

To examine the impact of stereotactic body radiation therapy (SBRT) dose on outcomes in early-stage non-small cell lung cancer in a large single-institution series.

Methods and Materials

We reviewed 600 patients treated from 2003 to 2012 for early-stage non-small cell lung cancer. The SBRT dose was at physician discretion on the basis of tumor size and location. Peripheral tumors were treated to 60 Gy in 3 fractions (homogeneous planning), 48-50 Gy in 4-5 fractions, or 30-34 Gy in 1 fraction. Central tumors were treated to 50 Gy in 5 fractions, 60 Gy in 8 fractions, or 50 Gy in 10 fractions. Patient, tumor, and treatment factors were assessed for their impact on patterns of failure, toxicity, and survival.

Results

An SBRT dose of 54-60 Gy in 3 fractions was associated with a statistically significant lower rate of local failure (LF) (4.3% at 2 years) compared with 30-34 Gy in 1 fraction (21%), 48-50 Gy in 4-5 fractions (15.5%), and 50-60 Gy in 8-10 fractions (13.3%). Lower pre-SBRT hemoglobin and higher positron emission tomography standardized uptake value were also associated with LF. Nodal failure, distant failure, and overall survival were similar between fractionation groups. Pulmonary toxicity (crude rate, any grade) was slightly higher for 3 fractions (5.0%) compared with 1 (3.2%) or 4-5 fractions (3.8%). Chest wall toxicity was also higher for 3 (23.7%) compared with 1 (8.6%) or 4-5 (7.7%) fraction regimens.

Conclusions

Although higher biologically equivalent dose SBRT (150-180 Gy10) may be associated with slightly lower LF, it was also associated with mildly increased toxicity and no difference in other patterns of failure or overall survival.

Introduction

Stereotactic body radiation therapy (SBRT) continues to mature as the treatment of choice for medically inoperable early-stage non-small cell lung cancer (NSCLC). A range of reported treatment regimens from 30 to 60 Gy in 1-8 fractions have all been described, with similar rates of local control (LC) 1, 2, 3, 4, 5, 6. Dose-fractionation regimens for SBRT are frequently selected at the discretion of the treating physician, utilizing a risk-adapted approach with regard to tumor size, location, and proximity to critical normal tissues.

In an effort to establish the ideal treatment regimen, early comparisons of available data suggested that the dose–response curve plateaued at a biologically equivalent dose (BED) of 100-105 Gy10 1, 7, 8, 9, 10 when using the linear-quadratic formula to adjust for SBRT dose (11). Although this threshold has been widely accepted, the initial Indiana University phase 1 dose-escalation study actually identified improved LC at higher BED levels (9 local failures at doses ≤16 Gy × 3, only 1 failure at doses 18-22 Gy × 3) (12), and more recent data have suggested that the dose–response curve may not be saturated at the classic BED of 100 Gy10, with improvement in LC (13) and overall survival (OS) (14) attainable at higher dose levels. Simultaneously, an international meta-analysis has suggested that very high BED may be associated with increased toxicity with a negative impact on outcomes, particularly in T2 tumors (15). Although Radiation Therapy Oncology Group (RTOG) protocol 0915, the only prospective randomized study reported to date, demonstrated similar outcomes in patients receiving 34 Gy in a single fraction compared with 48 Gy in 4 fractions (6), a proposed follow-up study comparing these regimens with the higher BED RTOG 0236 (2) has not yet been completed.

Given the limited availability of randomized data comparing standard regimens (BED ≥ 100 Gy10) to higher-dose regimens (BED ≥ 150 Gy10), we investigated the outcomes of these SBRT regimens within a large single-institution registry to compare differences in tumor control and toxicity among commonly used SBRT regimens.

Section snippets

Methods and Materials

We queried our institutional review board–approved prospective registry to identify patients with a new diagnosis of early-stage (T1-3 N0 M0) NSCLC treated with definitive intent SBRT between December 2003 and December 2012. Patients were immobilized in a Bodyfix (Elekta, Stockholm, Sweden) system with abdominal compression to restrict respiratory motion. In cases in which motion could not be adequately restricted to <1 cm, Active Breathing Control (Elekata) was used. Tumors within a 2-cm

Patient characteristics

Six-hundred three patients with 662 lesions (544 stage I, 59 synchronous primary NSCLC) were identified. Patient characteristics are presented in Table 1, with 14% receiving a single fraction (BED 120-149.6 Gy10), 21% receiving 3 fractions (BED 150-180 Gy10), 60% receiving 4-5 fractions (BED 100-105 Gy10), and 5% receiving 8-10 fractions (75-105 Gy10). Patient characteristics in this retrospective series were well matched for age, gender, performance status, Charlson comorbidity score, and

Discussion

This large single-institution series suggests that higher BED SBRT dosing to 60 Gy in 3 fractions (or the heterogeneity-corrected equivalent of 54 Gy in 3 fractions, BED 150-180 Gy10) leads to improved LC, however perhaps with a slightly higher rate of pulmonary and chest wall toxicity compared with regimens of 30-34 Gy in 1 fraction and 48-50 Gy in 4-5 fractions (BED 100-149.6 Gy10). Improved LC did not translate into differences in patterns of systemic failure (nodal or distant) and did not

Conclusion

Local control seems to be increased by high BED SBRT regimens ≥ 150 Gy10. Despite increased LC, there was no clear relationship between SBRT dose and other patterns of failure or OS. Further analysis is needed to determine whether a subset of patients with clear benefits from dose escalation can be identified. Ultimately, given the limitations of retrospective series, additional randomized trials comparing SBRT regimens with BED ranging from 100 to 180 Gy10 are warranted.

References (22)

Cited by (0)

This work was presented as an oral presentation at the 56th Annual Meeting of the American Society for Radiation Oncology, September 14-17, 2014, San Francisco, CA.

Conflict of interest: none.

View full text