Clinical Investigation
Stereotactic Ablative Radiation Therapy for Pulmonary Metastases: Histology, Dose, and Indication Matter

https://doi.org/10.1016/j.ijrobp.2017.02.093Get rights and content

Purpose

To assess the association between colorectal cancer (CRC) histology, dose, and local failure (LF) after stereotactic ablative radiation therapy (SABR) for pulmonary metastases, and to describe subsequent cancer progression, change of systemic therapy (CST), survival, and their association with treatment indications.

Methods and Materials

From a prospective SABR cohort, 180 pulmonary metastases in 120 patients were identified. Treatment indications were single metastasis, oligometastases, oligoprogression, and dominant areas of progression. Doses of 48 to 52 Gy/4 to 5 fractions were delivered. Since 2010 the dose for peripheral CRC metastases was increased to 60 Gy/4 fractions. Cumulative incidence function (CIF) was used to report LF, progression probability, and CST. The Kaplan-Meier method estimated overall survival (OS). Univariate and multivariable analyses to assess variable associations were conducted.

Results

Median follow-up was 22 months (interquartile range, 14-33 months). At 24 months, the CIF of LF was 23.6% (95% confidence interval [CI] 15.1%-33.3%) and 8.3% (95% CI 2.6%-18.6%), respectively, for CRC and non-CRC metastases (P<.001). This association remained significant after adjusting for confounders (subdistribution hazard ratio [SHR] 13.6, 95% CI 4.2-44.1, P<.001). Among CRC metastases, 56 and 45 received <60 Gy and 60 Gy, respectively. Delivering 60 Gy was independently associated with a lower hazard of LF (SHR 0.271, 95% CI 0.078-0.940, P=.040). At 12 months the CIF of progression was 41.67% (95% CI 21.69%-60.56%), 42.51% (95% CI 29.09%-55.29%), 62.96% (95% CI 41.25%-78.53%), and 78.57% (95% CI 42.20%-93.48%), respectively, for patients treated for single metastasis, oligometastases, oligoprogression, and dominant area of progression (P<.001). A CST was observed, respectively, in 4 (17%), 17 (31%), 12 (44%), and 10 (71%) patients with a median time of 13.1, 11.1, 8.4, and 8.4 months.

Conclusion

Colorectal cancer lung metastases are associated with a higher hazard of LF and require higher SABR doses. Outcomes for patients with oligometastases and oligoprogression treated with SABR seem favorable. Prospective clinical trials are needed to confirm these benefits.

Introduction

The management of patients with distant metastases predominantly involves systemic therapy (ST). The role of radiation in such patients was limited to a palliative one. However, that paradigm is changing. There is growing interest in considering high-dose radiation therapy and/or surgery to eradicate tumors, especially when the metastases are limited.

New high-precision radiation therapy techniques have allowed for noninvasive ablation of metastases (1). Stereotactic ablative radiation therapy (SABR) enables the delivery of high doses to the tumor while restricting the doses to adjacent normal tissues to a minimum (2). There is mounting evidence suggesting the safety and efficacy of SABR in the setting of pulmonary metastases, with 2-year control rates ranging from 70% to 90% 3, 4, 5, 6. Nevertheless, prognostic data from prospective trials are limited. Previous retrospective reports have suggested that pulmonary metastases from colorectal adenocarcinoma (CRC) might have a higher rate of local failure (LF) compared with other primary histology and that there is an association between the dose delivered and LF 7, 8.

The use of SABR to treat pulmonary metastases is increasing despite no randomized evidence, particularly in the treatment of “oligometastases” (9). The oligometastatic state, first described by Hellman and Weichselbaum (10), refers to a state between local-regional and widespread metastatic disease in which metastases are limited in number and location. In such a scenario, eradication of all sites of visible disease may result in prolonged progression-free survival and perhaps “cure” in a small proportion of patients, as shown in nonrandomized studies documenting the efficacy of resecting oligometastatic tumors 11, 12.

Another emerging indication for SABR is in the situation termed “oligoprogression.” This refers to the scenario in which cancer progression occurs in a limited number of tumors while the majority of other metastases are responding or stable while on an ST strategy. Rather than changing ST, one approach is to use a local therapy, like SABR, to treat the progressing tumors while staying on the same ST strategy (13). Such an approach has been reported in patients who develop oligoprogression while receiving targeted agents for non-small cell lung cancer and kidney cancer 14, 15, 16. The goal of such a strategy is to delay change of ST (CST).

This study aims to review the clinical outcomes of pulmonary metastases and identify predictors of success. The primary hypothesis is that among patients treated with SABR for pulmonary metastases, CRC primary is associated with a higher hazard of LF after adjusting for potential confounders. The secondary hypotheses are that: (1) dose is an independent predictor of LF among patients with metastases from CRC primary; and (2) there is an association between treatment indication and progression, CST, and overall survival (OS).

Section snippets

Population

Consecutive pulmonary metastases treated with SABR between November 2008 and December 2013 were identified from a prospective SABR cohort. This study was approved by our institutional research ethics board. Pulmonary metastases in adult patients with any solid primary tumor were included. Indications for SABR were as follows: (1) Single metastasis and oligometastases, for which the goal was to irradiate all sites of disease (≤5 active metastases; tumors previously ablated surgically or by

Patient and tumor characteristics

One hundred eighty-four consecutive pulmonary metastases in 120 patients were treated with SABR; 101 and 83, respectively, from CRC and non-CRC primary. The median follow-up was 22 months (IQR, 14-33 months). No patients were excluded in this analysis. Patient and tumor characteristics in the entire cohort and by primary site (CRC vs non-CRC) are described in Table 1. Overall they were equally distributed between the 2 groups, except for SABR indication: oligometastases was the most common SABR

Discussion

This study reported on the outcomes of SABR for pulmonary metastases from a large single institution. Colorectal cancer metastases were more likely to fail locally when compared with non-CRC, suggesting an underlying difference in tumor biology and radioresistance. This is in line with previous findings from observational studies 7, 20, 21, 22, 23. Takeda et al (7) reported a 1-year LC of 80% and 94%, respectively, for 21 CRC and 23 non-CRC pulmonary metastases treated with SABR. Similarly a

References (30)

Cited by (51)

  • Toxicity and Efficacy of Multitarget Thoracic Stereotactic Body Radiation Therapy

    2023, International Journal of Radiation Oncology Biology Physics
    Citation Excerpt :

    Assuming normal tissue constraints are achieved, this suggests that a single-isocenter approach is feasible in select cases and results in a modest increase in dose to the lungs. Colorectal histology was associated with a higher risk of LF after SBRT and, as indicated in multiple previous studies, this is thought to be due to the underlying differences in tumor biology and inherent radioresistance present in colorectal metastases.28-32 On univariate analysis, higher BED was found to be predictive of increased LF, which is contrary to the standard belief that higher dosing will improve local control.

  • Single-Fraction Stereotactic Ablative Body Radiotherapy to the Lung – The Knockout Punch

    2022, Clinical Oncology
    Citation Excerpt :

    A German series with 94 lung metastases treated with single-fraction SABR (17–26 Gy, median 24 Gy) or multi-fraction SABR (20–60 Gy, median 45 Gy in two to twelve fractions) reported higher 1- and 2-year local control of 89% and 83% with single-fraction SABR compared with 75% and 59% with multi-fraction SABR (P = 0.026) [25]. This finding should be interpreted with caution, as the lesions receiving single-fraction SABR were significantly smaller (median 12 mm versus 16 mm), more peripheral (71.1% versus 42.4%) and had fewer CRC primaries (17.8% versus 42.9%) [25,26]. Conversely, a Dutch study with 327 lung metastases (57.3% CRC, 17.5% NSCLC, 5.3% melanoma) showed 30 Gy single-fraction SABR (BED10 = 120 Gy) to be associated with lower local control (hazard ratio 3.63, P < 0.001) compared with multi-fraction SABR with BED10> 100 Gy in three to five fractions in unadjusted univariable analysis; this was not significant in multivariable analysis adjusted for factors including CRC histology and chemotherapy [27].

View all citing articles on Scopus

Conflict of interest: none.

View full text