Surgery, specifically a lobectomy or possibly pneumonectomy, has been the longtime historical standard of care for fit patients with early-stage, operable non-small cell lung cancer (NSCLC).  As focal radiation techniques have been refined, however, stereotactic body radiation therapy (SBRT), now alternatively known by the racier abbreviation SABR (sounds like "saber", stands for Stereotactic ABlative Radiation) has led to a growing body of evidence that this can lead to very good local control and also an encouraging overall survival (OS) for patients with stage I NSCLC who were considered "medically inoperable" due to medical issues that made surgery infeasible or far more dangerous than usual, though survival in such patients is very limited in large part due to their competing comorbidities. But as we have come to accept that SABR is a strong option for patients who cannot safely undergo surgery, patients and physicians have begun to ask the question of whether this approach might serve as a perfectly good alternative to surgery even for people who are fit enough to pursue it.

Most of the research addressing this question has come from series of patients at a single center with a clinical stage I NSCLC who are fit for surgery but declined to pursue it.  This work has largely supported the conclusion that SABR leads to good local control and OS rates that rival those of the patient to undergo surgical resection. In this setting, 3 different randomized clinical trials were launched several years ago to directly compare surgery to SABR in patients with smaller clinical stage I, resectable NSCLC. But because patients and their doctors tend to have feelings in favor of one approach or the other, it has been extremely hard to complete trials that randomize patients.  In fact, these trials have all closed after failing to enroll sufficient patients in a reasonable time. 

Is there any way to salvage any useful conclusions from these aborted trials?  Because two of them, known as  STARS and ROSEL, share very similar patient populations and trial designs, investigators pooled their results and published the preliminary findings, which are very provocative, albeit still with small numbers, even pooled together: SABR looks like a very strong alternative, if not preferable.

In both studies with stage I, node-negative NSCLC by pre-treatment workup were randomized to either SABR or lobectomy and mediastinal node sampling or dissection; the primary endpoint was OS for both trials. Even pooling results, there were only 58 patients (31 receiving SABR, 27 undergoing surgery), who were followed for a median of approximately three years after being treated.

The 3-year OS was 95%, with a single death, in the recipients of SABR, vs 79%, with six deaths, among surgery patients which was a significant difference (HR 0.14, log rank p = 0.037). Recurrence-free survival at 3 years was comparable whether patients underwent SABR or surgery, at 86% and 80%, respectively (HR 0.69, log-rank p = 0.54).

The side effect profile of SABR also looked encouraging. Only 3 patients (10%) experienced moderate or severe adverse effects, and none died from treatment. Among surgery patients,  there was a single death (4%) from operative complications, while 12 patients (44%) experienced moderate to severe treatment-related side effects.

Obviously, while these results suggest SABR is an attractive option, it would be a mistake to interpret the results from just a few dozen patients as a definitive answer. Importantly as well, we need longer follow up of these early stage patients, who are at risk for recurrences over a long period of many years.  It's certainly possible that patients on the surgery arm may fare better with longer follow-up.  And it's worth noting that patients who undergo SABR have essentially uninterpretable scan results around their treated lesion for years, as a "post-treatment scarring" effect leaves a hazy mess that requires us to scratch our heads for years before determining whether it will shrink, remain stable, or grow over time.

It may not be possible to ever answer this question definitively with a large, randomized head-to-head comparison of SABR vs. resection in fit patients with early stage disease, but that won't stop us from trying.  Additional trials are being developed through the VA system in the US, and also in the UK.  One potential modification is to have the trial discussed with patients, and consent obtained, by a “neutral party” such as a pulmonologist, research nurse, and/or multidisciplinary team, instead of by a thoracic surgeon or radiation oncologist who may convey bias that undermines the randomization.  

This work doesn't provide a definitive answer to whether SABR provides equivalent or even potentially superior OS in these operable, clinical stage I NSCLC patients, but we can at least say that these results corroborate the favorable findings in single institutional series with SABR. I think that even if SABR isn't a new standard of care, it is very fair to consider SABR as a reasonable alternative to surgery in early stage patients, even if they are candidates for surgery.