SABR Linked to Higher 3-Year Mortality vs Surgery in Early-Stage NSCLC

According to results from an emulated target trial published in the Journal of Clinical Oncology, adjusted 3-year all-cause mortality was higher after stereotactic ablative radiotherapy (SABR) than after surgical resection among patients with early-stage non–small cell lung cancer (NSCLC) who were eligible for either treatment.¹

The analysis of 12,546 patients in England estimated 3-year all-cause mortality at 30.3% with SABR vs 21.6% with surgery, a mean absolute risk difference of 8.8 percentage points (pp; 95% CI, 1.4-16.2; P = .02). A clinical panel had set the threshold for a clinically meaningful difference at 7.5 pp. Lung cancer–specific mortality was 20.6% vs 14.5% (6.1 pp; 95% CI, –0.6 to 12.9; P = .07), a difference that did not reach statistical significance.

Study Design

Investigators used England’s National Cancer Registry data, linked to Hospital Episode Statistics, the National Radiotherapy Data Set, and national death records. They identified patients diagnosed with T1 to -2N0M0 NSCLC from 2015 to 2020 who underwent SABR (n = 2299) or surgical resection (n = 10,247) within 6 months of diagnosis.¹ Eligible patients were aged 18 to 79 years with Zubrod performance status 0 to 2, or 80 years and older with a performance status of 0 to 1.

Patients who received SABR were older (median age, 74.9 vs 71.8 years) and more often had a performance status of 2 (26.0% vs 7.4%). They also had higher rates of chronic obstructive pulmonary disease (45.8% vs 26.0%) and ischemic heart disease (26.5% vs 16.5%). To reduce confounding by indication, the investigators used an instrumental variable analysis. The instrument was the historical proportion of eligible patients who received SABR rather than surgery within each referral network in the 6 months before treatment. The primary end point was 3-year all-cause mortality.

Mortality, Recurrence, and Readmission Outcomes

Unadjusted, 906 patients (39.4%) in the SABR group and 2033 (19.8%) in the surgery group had died by 3 years.¹ After adjustment, the differences were smaller at earlier time points: 7.4 pp at 2 years (95% CI, 0.9-13.8; P = .02) and 3.3 pp at 1 year (95% CI, –1.9 to 8.5; P = .21). The adjusted HR for death with SABR vs surgery was 1.94 (95% CI, 1.39-2.50) at 5 years.

Adjusted recurrence at 3 years was 34.3% with SABR vs 27.9% with surgery (6.5 pp; 95% CI, –1.7 to 14.6; P = .12). Adjusted 90-day readmissions were lower after SABR (23.3% vs 30.2%; –6.9 pp; 95% CI, –14.3 to 0.4; P = .07).¹

Findings in Older Patients

The adjusted 3-year mortality difference narrowed with age: 10.9 pp for patients 18 to 69 years (95% CI, 3.0-18.9), 8.4 pp for those 70 to 79 years (95% CI, 0.7-16.0), and 4.3 pp for those 80 years and older (95% CI, –3.7 to 12.3; P = .62).¹ All patients 80 years and older had performance status 0 to 1, and the difference in this group fell below the prespecified 7.5-pp threshold. Predicted 3-year all-cause mortality after surgery was 30.8% for patients 80 years and older vs 16.4% for those 18 to 69 years.

“The smaller difference compared with the overall population may reflect an increased risk of death after surgery in this age group,” the authors noted in the publication.

The difference also narrowed in more recent diagnosis years, from 12.4 pp in 2015 to 5.2 pp in 2020 (95% CI, –2.4 to 12.8). The authors said this trend “may indicate a learning curve associated with improvements in the technical quality of SABR delivery in England,” as case volume rose and new centers underwent accreditation and quality assurance.

Clinical Context and Limitations

Randomized comparisons of SABR and surgery have been challenging to enroll. ROSEL (NCT00687986), STARS (NCT02357992), SABRTooth (NCT02629458), and ACOSOG 4099/RTOG 1021 (NCT01336894) closed early because of slow accrual or were judged infeasible. Two ongoing phase 3 trials, VALOR (NCT02984761) and STABLEMATES (NCT02468024), are likely to underrepresent patients older than 80 years, according to the authors.

The authors contrasted their findings with a propensity-matched comparison of the revised STARS trial and video-assisted thoracic surgery in US practice, which reported 3-year overall survival of 91% in both groups.² In the current analysis, adjusted 3-year overall survival was 69.7% with SABR and 78.4% with surgery.¹ They said the gap may reflect healthier populations and more intensive follow-up in the earlier study.

Limitations include instrumental variable assumptions that cannot be tested directly and incomplete clinical staging in the registry. Competing risks from noncancer death may have led to underestimation of the effect of surgery on cancer-specific outcomes, and the data lacked patient-reported outcomes such as health-related quality of life. An inverse probability–weighted regression adjustment, which assumes no unmeasured confounding, produced a larger 3-year mortality difference of 14.5 pp (95% CI, 11.6-17.4).

The authors concluded that surgical resection should remain the standard primary treatment, while SABR is a curative-intent alternative, particularly for older patients.¹

REFERENCES

1. Kagenaar E, Lugo-Palacios DG, Aggarwal A, et al. Stereotactic ablative radiotherapy versus surgical resection for early-stage non–small cell lung cancer: an emulated target trial. J Clin Oncol. Published online November 2, 2026. doi:10.1200/JCO-26-00399

2. Chang JY, Mehran RJ, Feng L, et al. Stereotactic ablative radiotherapy for operable stage I non-small-cell lung cancer (revised STARS): long-term results of a single-arm, prospective trial with prespecified comparison to surgery. Lancet Oncol. 2021;22:1448-1457.