A phase III comparison of prophylactic cranial irradiation versus observation in patients with locally advanced non-small cell lung cancer: Initial analysis of Radiation Therapy Oncology Group 0214

Reviewer: Arpi Thukral, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 3, 2009

Presenter: Elizabeth M. Gore, MD
Presenter's Affiliation: Medical College of Wisconsin
Type of Session: Scientific


  • Central nervous system (CNS) failure rates are high in patients with locally advanced non-small cell lung cancer.
    • As our therapies for NSCLC improve and loco-regional control and survival are increased in this population, rates of CNS metastases in this group are increasing as well. The brain is often the only site of relapse in these patients.
    • Current survival rates for patients with locally advanced NSCLC are up to 20 months with 30-55% CNS failure rates overall.
  • Although many have examined therapies for reducing CN metastases risk in the past, there is currently no standard agreement as to how to appropriate address this risk.
  • Prophylactic cranial irradiation (PCI) has been used for many years in patients with limited-stage small cell lung cancer. Although small cell lung cancer usually responds very well to chemotherapy, most agents which are used do not cross the blood-brain barrier and therefore, the likelihood brain metastases after completion of therapy can be as high as 50% in patients who live for two years after diagnosis.
    • The use of prophylactic cranial irradiation (PCI) in patients with limited stage small cell lung cancer (SCLC) has been shown to provide an overall survival benefit (3-year overall survival of 20.7% with PCI versus 15.3% without) in a large meta-analysis by Auperin et. al. (NEJM, 1999) by decreasing the risk of developing brain metastasis. For this reason, PCI has become the standard of care for patients with limited-stage SCLC. A recent study by Slotman, et. al. (NEJM, 2007) has shown that patients with extensive stage small cell lung cancer who had a response to chemotherapy may also have a survival benefit.
  • The concept of PCI was introduced for non-small cell lung cancer (NSCLC) in the 1980’s by various groups including VALG, RTOG, and MDACC, however none of these trials found a statistically significant survival difference.
  • Since many of these studies were done over 20 years ago, it was thought by the RTOG that with new treatment modalities and options for patients with NSCLC, we may see a survival advantage with PCI in the current era.
  • The purpose of the current trial was to perform a phase III trial comparing prophylactic cranial irradiation versus observation in patients with locally advanced non-small cell lung cancer.  This study was conducted by the Radiation Therapy Oncology Group.

Materials and Methods

  • Eligibility criteria:
    • Patients with stage IIIA or IIIB NSCLC without progression of disease after loco-regional treatment with surgery and/or radiation therapy with or without chemotherapy.
    • They had to complete loco-regional treatment within 16 weeks of starting the trial.
    • They had to have a complete response (CR), partial response (PR), or stable disease (SD) after initial loco-regional treatment.
    • Acute or sub-acute toxicities to loco-regional treatment < Grade 2.
    • No CNS metastases by MRI brain or CT head.
  • Participants were randomized to prophylactic cranial irradiation (PCI) or observation and stratified by stage (IIIA or B), histology (non-squamous or squamous) and therapy (surgery or no surgery).
  • PCI was delivered once daily at 2 Gy per fraction to 30Gy in the experimental group.
  • The primary endpoint of the study was overall survival (OS).
  • Secondary endpoints were disease free survival (DFS) and the impact of PCI on incidence of CNS metastases, neuropsychological function, and quality of life (QoL).
  • Kaplan- Meier estimation with the log-rank test was used for OS and DFS and the logistic regression model was used for calculating the incidence of CNS metastasis.
  • Target accrual for the study was 1058 patients, however the study was closed early due to slow accrual, and only 356 patients total were accrued.


  • Total accrual was 356 patients between 9/19/02 and 8/30/07.
  • 340 patients of these patients were evaluable.
  • Patient characteristics were similar between the 2 arms.
    • Median age of all patients was 63.
    • Two-thirds of patients were male and the Zubrod Performance Status for the majority of patients was 0-1.
    • One-third of patients had squamous cell histology.
    • The distribution of IIIA and IIIB lung cancers were equal between the 2 groups.
    • Two-thirds of patients had no surgery.
  • One year OS: 75.6 % and 76.9% for PCI and observation, respectively; p=0.86: NS.
  • One year DFS: 56.4% and 51.2% for PCI and observation, respectively, p=0.11: NS .
  • However, the CNS metastases rate at 1 year was statistically significantly different with CNS relapse 7.7% vs. 18% for PCI vs. observation; p=0.004.
  • Logistic regression showed that the patients in the observation arm are 2.52 times more likely to develop CNS metastases than those in the PCI arm (odds ratio=2.52, 95% CI=(1.32-4.80)).
  • Grade I and II acute toxicities seen were mostly fatigue and nausea. Grade I and II late toxicities included fatigue, mild headaches, and alopecia.
    • 4 patients had Grade III toxicities which included syncope, dyspnea, and weakness.

Author's Conclusions

  • PCI in patients without progressive disease after loco-regional therapy for III NSCLC significantly decreases the rate of CNS metastases.
  • There was no statistically significant difference in OS or DFS seen for patients receiving PCI.
  • Acute and late toxicities were not severe.
  • Analyses of the impact of PCI on neuropsychological function and QoL will be presented at the World Lung Conference this summer.
  • The authors conclude that additional follow-up is needed prior to changing our current practice patterns regarding PCI in advanced NSCLC patients.

Clinical/Scientific Implications

  • A previous study by Pottgen et al. published in JCO 2007 showed that in patients with Stage IIIA NSCLC, PCI was effective in preventing brain metastases following aggressive an trimodality approach of treatment, including surgery, XRT, and chemotherapy. Neurocognitive late effects were not significantly different between patients treated with or without PCI.
  • Based on the results of this study and previous trials done on PCI in NSCLC, the RTOG 0214 study was well-designed with a rational hypothesis, however it failed to meet its primary endpoint.
  • Although, there was a decrease in rates of CNS metastases seen in the group who received PCI, there was no difference seen in OS over the 1st year of follow up for these patients with the addition of PCI.
  • Reasons for this may be:
    • There was a very low accrual in this study from what was planned. The investigators, despite their valiant effort, were only able to accrue one-third of the planned patients over a 5 year time period. It is unclear if we would have seen a survival difference if the numbers of patients on this study met the power need to support an OS difference.
    • Longer-term follow may be needed to see and OS difference.
    • Data from the QoL and neurocognitive studies will be important to see, as this may have an impact on whether or not PCI may still be beneficial to patients, despite not having a survival differences.
  • At this time, there is no role for use of PCI in the NSCLC population. We will await the results of QOL data to be presented later.

See Patient Summary