Nomogram Predicts Metastases After SBRT in Non-Small Cell Lung Cancer

For patients with inoperable early-stage non-small cell lung cancer (NSCLC), stereotactic body radiation therapy (SBRT) is the standard of care. High doses of radiation to a precisely targeted area typically provide good tumor control locally. However, metastases often develop in the brain, bone, liver and other areas of the body.

While chemotherapy can help control distant disease, many patients too sick for surgery — due to heart disease, COPD or other comorbidities — are often too sick for chemotherapy as well, says Aditya Juloori, MD, a radiation oncology resident working with Kevin Stephans, MD, at Cleveland Clinic Cancer Center.

“The advent of immunotherapy has changed the outlook for these patients, offering a better tolerated treatment option for distant disease,” he says. “That has renewed interest in identifying patients more likely to develop metastases. Which patients should be considered for immunotherapy following SBRT?”

Dr. Juloori, Dr. Stephans and colleagues at Cleveland Clinic collaborated with Washington University School of Medicine to develop a predictive nomogram revealing the risk of metastases within one year of SBRT for NSCLC. It will be introduced in an oral presentation at the 2018 ASTRO Annual Meeting in San Antonio.

“Our model incorporates various patient and tumor criteria,” says Dr. Stephans, “but we found the two most important factors in predicting metastatic disease to be tumor size and PET standardized uptake value (SUV). The larger the tumor and higher the SUV, the higher the risk of distant metastases after SBRT.”

Risk of metastases from 10 to 70 percent

The records of approximately 1,000 early-stage NSCLC patients treated with lung SBRT at Cleveland Clinic were used to develop the model, using random forest, regression analysis and other statistical modeling tools. Another set of more than 700 patients from Washington University was used to validate it.

Median overall survival of the first and second cohorts was 1.71 and 1.92 years, respectively. Median follow-up was 18.3 months and 21.1 months. One-year incidence of distant failure was 16 and 12.1 percent.

Of all patients with tumor size >3 cm, incidence of distant metastases at one year was 26 percent. In those with tumor size ≤3 cm, incidence was 12.6 percent.

Of all patients with PET SUV ≥4.1, incidence of distant metastases at one year was 18.5 percent. In those with PET SUV <4.1, incidence was 8.4 percent.

The final nomogram includes tumor size, histology, PET SUV, age, Karnofsky Performance Scale (KPS) score, and active smoking status, and has a cross-validated C-index of 0.62. The nomogram predicts the likelihood of distant metastases at one year between 10 and 70 percent.

Guiding patient selection for immunotherapy

“Until now, we haven’t had a good understanding of what patient and tumor characteristics cause SBRT to fail in patients with early-stage NSCLC,” says Dr. Juloori. “Understanding these factors — as well as using the nomogram to identify patients at high risk for metastases — will help guide patient selection for potential inclusion in immunotherapy trials.”

More trials are needed to test both safety and efficacy of immunotherapy after SBRT, he notes.