A ‘Fool’s Errand’? Picking a Winner for Treating Early NSCLC

For years, the default definitive treatment for patients with early-stage I non–small cell lung cancer (NSCLC) has been surgical resection, typically minimally invasive lobectomy with systematic lymph node dissection.

Guidelines from the National Comprehensive Cancer Network (NCCN), the American Society of Clinical Oncology, and the European Society for Medical Oncology all list surgery (in particular, lobectomy) as the primary local therapy for fit, operable patients with stage I NSCLC.

More recently, however, stereotactic body radiotherapy (SBRT), also called stereotactic ablative radiotherapy, has emerged as a definitive treatment option for stage I NSCLC, especially for older, less fit patients who are unsuitable or deemed high-risk for surgery.

“We see patients in our practice who cannot undergo surgery or who may not have adequate lung function to be able to tolerate surgery, and for these patients who are medically inoperable or surgically unresectable, radiation therapy may be a reasonable option,” Charu Aggarwal, MD, MPH, professor and lung cancer specialist, University of Pennsylvania, Philadelphia, told Medscape Medical News.

Given some encouraging data suggesting that SBRT could provide similar survival outcomes and be an alternative to surgery for operable disease, SBRT is also increasingly being considered in these early-stage patients. However, other evidence indicates that SBRT may be associated with higher rates of regional and distant recurrences and worse long-term survival, particularly in operable patients.

What may ultimately matter most is carefully selecting operable patients who undergo SBRT.

Aggarwal has encountered certain patients who are fit for surgery but would rather have radiation therapy. “This is an individual decision, and these patients are usually discussed at tumor board and in multidisciplinary discussions to really make sure that they’re making the right decision for themselves,” she explained.

The Pros and Cons of SBRT

SBRT is a nonsurgical approach in which precision high-dose radiation is delivered in just a few fractions — typically, 3, 5, or 8, depending on institutional protocols and tumor characteristics.

SBRT is performed on an outpatient basis, usually over 1-2 weeks, with most patients able to resume usual activities with minimal to no delay. Surgery, on the other hand, requires a hospital stay and takes most people about 2-6 weeks to return to regular activities. SBRT also avoids anesthesia and surgical incisions, both of which come with risks.

The data on SBRT in early-stage NSCLC are mixed. While some studies indicate that SBRT comes with promising survival outcomes, other research has reported worse survival and recurrence rates.

One potential reason for higher recurrence rates with SBRT is the lack of pathologic nodal staging, which only happens after surgery, as well as lower rates of nodal evaluation with endobronchial ultrasound or mediastinoscopy before surgery or SBRT. Without nodal assessments, clinicians may miss a more aggressive histology or more advanced nodal stage, which would go undertreated if patients received SBRT.

Latest Data in Large Cohort

A recent study published in Lung Cancer indicated that, when carefully selected, operable patients with early NSCLC have comparable survival with lobectomy or SBRT.

In the study, Dutch researchers took an in-depth look at survival and recurrence patterns in a retrospective cohort study of 2183 patients with clinical stage I NSCLC treated with minimally invasive lobectomy or SBRT. The study includes one of the largest cohorts to date, with robust data collection on baseline characteristics, comorbidities, tumor size, performance status, and follow-up.

Patients receiving SBRT were typically older (median age, 74 vs 67 years), had higher comorbidity burdens (Charlson index ≥ 5 in 57% of SBRT patients vs 23% of surgical patients), worse performance status, and lower lung function. To adjust for these differences, the researchers used propensity score weighting so the SBRT group’s baseline characteristics were comparable with those in the surgery group.

The surgery cohort had more invasive nodal evaluation: 21% underwent endobronchial ultrasound or mediastinoscopy vs only 12% in the SBRT group. The vast majority in both groups had PET-CT staging, reflecting modern imaging-based workups.

While 5-year local recurrence rates between the two groups were similar (13.1% for SBRT vs 12.1% for surgery), the 5-year regional recurrence rate was significantly higher after SBRT than lobectomy (18.1% vs 14.2%; hazard ratio [HR], 0.74), as was the distant metastasis rate (26.2% vs 20.2%; HR, 0.72).

Mortality at 30 days was higher after surgery than SBRT (1.0% vs 0.2%). And in the unadjusted analysis, 5-year overall survival was significantly better with lobectomy than SBRT (70.2% vs 40.3%).

However, when the analysis only included patients with similar baseline characteristics, overall survival was no longer significantly different in the two groups (HR, 0.89; 95% CI, 0.65-1.20). Lung cancer–specific mortality was also not significantly different between the two treatments (HR, 1.08), but the study was underpowered to detect significant differences in this outcome on the basis of a relatively low number of deaths from NSCLC.

Still, even after comparing similar patients, recurrence-free survival was notably better with surgery (HR, 0.70), due to fewer regional recurrences and distant metastases. Overall, 13% of the surgical cohort had nodal upstaging at pathology, meaning that even in clinically “node-negative” stage I disease, a subset of patients had unsuspected nodal involvement.

Patients receiving SBRT did not have pathologic nodal staging, raising the possibility of occult micrometastases. The authors noted that the proportion of SBRT patients with occult lymph node metastases is likely at least equal to that in the surgery group, but these metastases would go undetected without pathologic assessment.

Missing potential occult micrometastases in the SBRT group likely contributed to higher regional recurrence rates over time. By improving nodal staging, more patients with occult lymph node metastases who would be undertreated with SBRT may be identified before treatment, the authors said.

What Do Experts Say?

So, is SBRT an option for patients with stage I NSCLC?

Opinions vary.

“If you got one shot for a cure, then you want to do the surgery because that’s what results in a cure,” said Raja Flores, MD, chairman of Thoracic Surgery, Mount Sinai Health System, New York City.

Flores noted that the survival rate with surgery is high in this population. “There’s really nothing out there that can compare,” he said.

In his view, surgery “remains the gold standard.” However, “radiation could be considered in nonsurgical candidates,” he said.

The most recent NCCN guidelines align with Flores’ take. The guidelines say that SBRT is indicated for stage IA-IIA (N0) NSCLC in patients who are deemed “medically inoperable, high surgical risk as determined by thoracic surgeon, and those who decline surgery after thoracic surgical consultation.”

Clifford G. Robinson, MD, agreed. “In the United States, we largely treat patients with SBRT who are medically inoperable or high-risk operable and a much smaller proportion who decline surgery,” said Robinson, professor of radiation oncology and chief of SBRT at Washington University in St. Louis, St. Louis. “Many patients who are deemed operable are not offered SBRT.”

Still, for Robinson, determining which patients are best suited for surgery or SBRT remains unclear.

“Retrospective comparisons are fraught with problems because of confounding,” Robinson told Medscape Medical News. “That is, the healthier patients get surgery, and the less healthy ones get SBRT. No manner of fancy statistical manipulation can remove that fact.”

In fact, a recent meta-analysis found that the most significant variable predicting whether surgery or SBRT was superior in retrospective studies was whether the author was a surgeon or radiation oncologist.

Robinson noted that multiple randomized trials have attempted to randomize patients with medically operable early-stage NSCLC to surgery or SBRT and failed to accrue, largely due to patients’ “understandable unwillingness to be randomized between operative vs nonoperative interventions when most already prefer one or the other approach.”

Flores highlighted another point of caution about interpreting trial results: Not all early-stage NSCLC behaves similarly. “Some are slow-growing ‘turtles,’ and others are aggressive ‘rabbits’ — and the turtles are usually the ones that have been included in these radiotherapy trials, and that’s the danger,” he said.

While medical operability is the primary factor for deciding the treatment modality for early-stage NSCLC, there are other more subtle factors that can play into the decision.

These include prior surgery or radiotherapy to the chest, prior cancers, and social issues, such as the patient being a primary caregiver for another person and job insecurity, that might make recovery from surgery more challenging. And in rare instances, a patient may be medically fit to undergo surgery, but the cancer is technically challenging to resect due to anatomic issues or prior surgery to the chest, Robinson added.

A Winner?

Results from two ongoing, highly anticipated randomized trials expected in the next several years will hopefully provide additional insights and clarify ongoing uncertainties about the optimal treatment strategies for operable patients with stage I NSCLC.

STABLE-MATES is comparing outcomes after sublobar resection and SBRT in high-risk operable stage I NSCLC, and VALOR is evaluating outcomes after anatomic pulmonary resections and SBRT in patients with stage I NSCLC who have a long life expectancy and are fit enough to tolerate surgery.

But Robinson said his group believes that trying to decide on a winner is a “fool’s errand” and is instead running a pragmatic study across multiple academic and community centers around the United States and Canada where patients choose therapy based on their personal preferences and guidance from their physicians. The researchers will carefully track baseline comorbidity and frailty and assess serial quality-of-life changes over time.

“The goal is to create a calculator that a given patient might use in the future to determine what patients like them would have received, complete with expected outcomes and side effects,” Robinson said.

Robinson cautioned, however, that it “seems unlikely, given the existing literature, that one of the treatments will be truly ‘superior’ to the other one and lead to the ‘losing’ treatment fading away since both are excellent options with pros and cons.”

Aggarwal, Robinson, and Flores had no relevant disclosures.