Fabio Ynoe de Moraes, Associate Professor at Queen’s University, shared on LinkedIn:
“Executive takeaways
- MAVERICK is likely practice changing: MRI surveillance without PCI produced substantially better cognitive failure–free survival in SCLC, with no preliminary OS disadvantage. Final OS remains immature, and omission of PCI requires reliable serial MRI and prompt salvage treatment.
- SWOG/NRG S1914 is definitively negative: adding peri-SBRT atezolizumab to SABR for high-risk, medically inoperable early-stage NSCLC did not improve survival and increased serious toxicity, including two fatal respiratory events.
- No comparably important new evidence changed glioma, brain-metastasis SRS, spine SBRT, or oligometastatic SABR practice.
1. MAVERICK: MRI surveillance can replace PCI for many patients with SCLC
Evidence status: International randomized phase III trial; conference presentation only. Practice signal: Potentially practice changing, with final OS and peer-reviewed publication pending.
Clinical question
Among patients with limited- or extensive-stage SCLC who complete initial therapy without brain metastases, does structured MRI surveillance preserve cognition without compromising survival compared with PCI plus MRI surveillance?
Design and population
SWOG S1827/MAVERICK randomized 304 patients:
- Limited-stage SCLC: 68%
- Extensive-stage SCLC: 32%
- Upfront immunotherapy: 41%
- No brain metastases on post-treatment MRI
- MRI surveillance alone versus PCI plus identical MRI surveillance
- MRI every three months during year 1 and every six months during year 2
- Standardized serial cognitive testing
In the PCI arm, 77% received hippocampal-avoidance PCI. The original primary endpoint was OS, but slow accrual prompted a protocol amendment making cognitive failure–free survival the primary endpoint.
Results
OutcomeMRI alone versus MRI + PCICognitive failure–free survivalHR 0.6090% CI0.46–0.78PP value0.0005Estimated 6-month CFFS37.8% vs 16.5%Preliminary overall survivalHR 0.9090% CI0.67–1.20Grade ≥2 treatment-related toxicity7.1% vs 41.3%Grade ≥3 treatment-related toxicity0.8% vs 7.9%
The cognitive benefit was consistent across limited- and extensive-stage disease and was not materially modified by receipt of immunotherapy. PFS did not differ significantly.
PCI continued to alter the failure pattern: in the MRI-alone group, the brain was the first progression site in approximately 18%. Brain-metastasis–free survival numerically favored PCI but was not statistically different.
The OS analysis remains preliminary after 128 deaths; the final analysis is planned after 190 deaths. SWOG’s official MAVERICK results and IASLC/ILCN detailed report
Critical appraisal
Strengths
- Randomized phase III evidence
- Included both limited- and extensive-stage disease
- Contemporary MRI surveillance
- Prospective standardized cognitive testing
- Most PCI patients received hippocampal avoidance, making the comparison clinically contemporary
Limitations
- Primary endpoint changed from OS to cognitive failure–free survival.
- OS is immature and reported with a 90%, rather than conventional 95%, confidence interval.
- Twenty-nine patients assigned to PCI did not receive it, diluting an intention-to-treat comparison.
- Cognitive testing was limited to English- and French-speaking patients.
- ‘Cognitive failure’ was defined as decline on any one of six tests, producing low absolute six-month CFFS rates in both groups and making the endpoint sensitive to measurement variability.
- Quality-of-life and detailed domain-specific cognitive results remain incomplete.
- Trial-level surveillance depends on MRI access and timely salvage SRS/WBRT, limiting transportability to resource-constrained settings.
Practice interpretation
MAVERICK supports MRI surveillance as the preferred strategy for many informed patients, particularly those prioritizing cognition or at increased neurotoxicity risk.
I would now discuss omission of PCI when all of the following are present:
- High-quality baseline brain MRI
- Commitment to MRI every three months in year 1 and every six months in year 2
- Rapid review of surveillance imaging
- Ready access to salvage SRS or WBRT
- Patient understanding that intracranial relapse remains more likely without PCI
PCI remains reasonable when MRI surveillance or salvage treatment cannot be reliably delivered, and potentially for patients who strongly prioritize reducing CNS recurrence over cognitive risk.
My position: This changes the default discussion immediately, but definitive abandonment of PCI—especially in fit limited-stage patients—should await mature OS, full cognitive/QoL analyses, and peer-reviewed publication.
2. SWOG/NRG S1914: peri-SBRT atezolizumab should not be used in early-stage NSCLC
Evidence status: Peer-reviewed randomized phase III trial published in The Lancet. Practice signal: Definitive negative result.
Clinical question
Does adding induction, concurrent, and consolidation atezolizumab to SBRT improve overall survival in medically inoperable, high-risk early-stage NSCLC?
Design and population
- Multicentre, open-label randomized phase III trial
- 146 US institutions
- T1–T3N0M0 NSCLC, ≤7 cm
- Medically inoperable or declined surgery
- At least one recurrence-risk feature
- SBRT in 3–8 fractions
- Comparator: identical SBRT alone
- Experimental arm: atezolizumab 1200 mg every 21 days for up to eight cycles, with SBRT beginning during cycle 3
Primary endpoint: OS - A total of 417 patients were randomized; 402 were eligible for the modified intention-to-treat analysis. Median age was 72.8 years.
Results
Accrual stopped at the first interim analysis for futility. With median follow-up of 24.8 months among surviving patients:
OutcomeAtezolizumab + SBRT vs SBRTOverall-survival HR1.0495% CI0.69–1.58One-sided PP value0.58Estimated 2-year OS82% vs 82%Grade ≥3 adverse events12% vs 3%
Two grade 5 respiratory events occurred in the atezolizumab-SBRT group.
S1914 primary publication
Critical appraisal
This is a strong negative result: cooperative-group phase III design, broad multicentre participation, clinically relevant control arm and no suggestion of OS benefit.
The premature futility closure reduced statistical precision, as reflected by the wide OS confidence interval. However, the combination produced no efficacy signal sufficient to offset its clear toxicity increment.
The study does not prove that every immunotherapy-SABR strategy will fail. Possible determinants include:
- PD-L1 agent
- Patient selection and tumor biology
- Timing relative to radiation
- Treatment duration
- Competing mortality in an older, medically inoperable population
- Whether distant-recurrence risk is high enough to justify systemic therapyNevertheless, another trial would need compelling biomarker or mechanistic justification—not merely substitution of a different checkpoint inhibitor.
Practice interpretation
Do not add atezolizumab before, during, or after SBRT for medically inoperable early-stage NSCLC outside a clinical trial.
For standard practice:
- SBRT alone remains the curative standard for medically inoperable node-negative early-stage disease.
- There is no basis for off-label peri-SBRT atezolizumab.
- Patients receiving immunotherapy for another indication require careful assessment of pulmonary toxicity when thoracic SBRT is planned.
This result is particularly important strategically: immunotherapy does not automatically increase the value of thoracic radiation. Integration must be supported by randomized evidence, biological selection and clinically meaningful endpoints.”
Other articles about NSCLC on OncoDaily.