Phase III Study of Prophylactic Cranial Irradiation vs. Observation in Patients with Stage III Non–small-cell Lung Cancer: Neurocognitive and Quality of Life Analysis of RTOG 0214

Reviewer: Charles B. Simone, II, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: November 6, 2009

Presenter: B. Movsas
Presenter's Affiliation: Presenter's Affiliation: Henry Ford Hospital, Detroit, MI
Type of Session: Plenary


  • Fifty percent of patients with locally advanced non-small cell lung cancer (NSCLC) develop central nervous system (CNS) metastases at some time during the course of their disease, with the CNS as first site of relapse occurring in 15-30% of patients.
  • Prevention of CNS metastases may improve patient quality of life and prolong overall survival.
  • Although the addition of chemotherapy to thoracic radiation therapy reduces the risk of extracranial distant metastases and improves survival for patients with locally advanced NSCLC, chemotherapy does not alter brain relapse rates. Therefore, other treatment options directed at CNS micrometastases need to be considered.
  • Following the recent publications of a survival benefit in multiple randomized clinical trials and meta-analyses, prophylactic cranial irradiation (PCI) is increasingly being considered as the standard of care for patients with limited-stage small cell lung cancer (SCLC) with a complete or strong partial response to thoracic therapy and those with extensive-stage SCLC with an initial response to chemotherapy.
  • Up to half of patients with SCLC treated with concurrent chemotherapy and PCI have been found to have abnormal neuropsychologic tests. However, a significant change in those deficits after PCI has not definitively demonstrated, suggesting that neuropsychologic abnormalities associated with SCLC may be secondary to the disease itself. Overall, PCI has been demonstrated to have a favorable impact on quality of life among patients with SCLC.
  • The effects of PCI on neurocognitive function and quality of life in patients with NSCLC prospectively enrolled on a clinical trial have not previously been reported, and limited neuropsychologic testing has been conducted in this patient population.
  • Three randomized trials of PCI in patients with locally advanced NSCLC have previously been published. These studies collectively have demonstrated that PCI decreases or delays the incidence of brain metastases, although no survival benefit has been demonstrated.
  • Following the initial report of the primary endpoint of RTOG 0214, in which no overall survival benefit for PCI was demonstrated for patients with Stage III NSCLC, the current study assesses the secondary objectives in determining the impact of PCI on the incidence of CNS metastasis, neurocognitive function, and quality of life.

Materials and Methods

  • Eligibility criteria for the study included patients with newly diagnosed Stage IIIA or IIIB NSCLC that completed definitive locoregional therapy (surgery and/or radiation therapy, with or without chemotherapy) with complete response, partial response, or stable disease after therapy. Patients needed to be at least 18 years of age and enrolled within 16 weeks of completing previous therapy. An MRI or CT of the head showing no suspicion for CNS metastases within 6 weeks of study entry, no evidence of extracranial distant metastatic disease, and no prior cranial irradiation were also required.
  • Patients were randomized to PCI or observation. Radiation therapy was delivered in 2 Gy fractions to 30 Gy over three weeks.
  • Mini-mental status exam (MMSE), activities of daily living scale (ADLS), and Hopkins Verbal Learning Test (HVLT) were used to assess neurocognitive function to determine the neuropsychologic impact of the development of CNS metastases and the use of PCI. The EORTC QLQC30 and QLQBN20 were used to assess quality of life.
  • Follow-up assessment of patients included an MRI or CT of the brain at six and 12 months after study entry and then annually. MMSE, ADLS, and HVLT were conducted at 3, 6, 12, 18, 24, 30, 36, and 48 months after study entry. EORTC QLQC30 and QLQBN20 were conducted at 6, 12, 24, 36, and 48 months after study entry. Evaluation of locoregional disease was practitioner dependent, with chest imaging required at least every six months.
  • The cutoff of neurologic deterioration evaluated by MMSE and HVLT was calculated using reliable change index at 1 year from baseline. For QOL, a 10 point (of 100) change was considered clinically relevant.


  • The study was open from September 2002 to August 2007 and closed early due to poor accrual. A total of 356 patients of a planned 1,058 patients were accrued to this study, with 340 patients eligible for analysis.
  • There was no difference in 1-year overall survival (75.6% vs. 76.9%, p = 0.86) or 1-year disease free survival (56.4% vs. 51.2%, p = 0.11) for patients receiving PCI vs. observation, respectively.
  • The incidence of CNS metastasis at one year, however, was significantly lower in patients receiving PCI (7.7% vs. 18.0%, p = 0.004).
  • Among quality of life testing, at one year, there were no significant differences between the two arms in any component of the EORTC QLQC30 or QLQBN20 (p values for all parameters > 0.05), although a trend for greater decline in patient-reported cognitive functioning was noted among patients receiving PCI.
  • Among neurocognitive function testing, there was greater decline in immediate recall (p = 0.03) and delayed recall (p = 0.008) in the PCI arm at one year demonstrated with the HVLT. These differences were independent of patient age or the development of CNS metastasis. More patients in the PCI arm reported cognitive decline at 3 months (35% vs. 18%). There were no differences in MMSE (p = 0.60) or ADLS (p = 0.88).

Author's Conclusions

  • The use of PCI in patients with Stage III NSCLC significantly decreases the risk of CNS metastasis.
  • No significant improvement was demonstrated in overall survival or disease-free survival with the use of PCI.
  • There were no significant differences in global cognitive function based on MMSE testing or in quality of life following PCI. There was, however, a significant decline in memory, as assessed by the HVLT.
  • Despite its low accrual, this study provides prospective data regarding the relative benefits and risks of PCI for the treatment of patients with stage III NSCLC following definitive local therapy and reinforces the need to use sensitive cognitive assessments.

Clinical/Scientific Implications

  • This study represents the largest trial of PCI use among patients with NSCLC, and it is the first to assess quality of life and neurocognitive function prospectively using previously validated tools. Formal neuropsychometric evaluation remains the most comprehensive tool to assess cognitive and psychosocial function. Neuropsychologic function testing has been used successfully in clinical trials of agents for primary brain tumors and is becoming an increasingly important component of any trial administering irradiation to the brain.
  • When examining the tests used in this study to assess neuropsychological status, the MMSE is easy to administer and have been validated previously for mild dementia. Although it has a high specificity, it lacks sensitivity. The HVLT is a well-validated and reliable assessment of memory over time, with an improved sensitivity compared with the MMSE. ADLS provides vital information on day-today quality of life that is not covered by MMSE, HVLT, or physical examination.
  • When examining the tests used in this study to assess quality of life, The EORTC QLQC30 is a 30-item, self report questionnaire assessing physical functioning, role functioning, emotional functioning, cognitive functioning, social functioning, global quality of life, fatigue, pain, nausea and vomiting, dyspnea, insomnia, anorexia, constipation, diarrhea, and financial impact. QLQBN20 is a 20-item assessment of future uncertainty, visual disorder, motor dysfunction, and communication deficit, and it includes an assessment of headache, seizure, drowsiness, hair loss, itching, weakness of both legs, and difficulties with bladder control.
  • With the use of multiple tests to assess both neuropsychological status and quality of life, the authors should be commended for such an extensive evaluation of these very important patient-centered symptoms.
  • However, several limitations to this study exist. Not all patients underwent baseline or follow-up MRI neuroimaging. It is possible that some patients enrolled in this trial with no evidence of metastasis on CT head imaging would have had intracranial metastatic disease had they been screened with and MRI brain. Additionally, the low accrual in this study might have prevented the detection of a small, but statistically significant, benefit in overall survival with PCI in this population on the magnitude of approximately 5%, as has been demonstrated with SCLC. The low accrual also will not allow for subgroup analysis to assess if PCI is more beneficial for certain histologic subtypes of NSCLC that may be more likely to result in CNS metastasis. Additionally, although the decline in compliance with neurocognitive testing with time was less than that reported in other similar PCI trials, only 144 patients at 6 months and 95 patients at 12 months completed this testing. This may have limited the authors’ ability to detect a change in quality of life or neurocognitive functioning that may truly exist following PCI. Means to limit this compliance decline with time are currently being assessed in RTOG 0828, “Pilot Project To Reduce Missing RTOG Quality Of Life Data Via Electronic Web-Based Form Collection: A Companion Study For RTOG 0415 (A Phase III Randomized Study Of Hypofractionated 3D-CRT/IMRT Versus Conventionally Fractionated 3D-CRT/IMRT In Patients With Favorable-Risk Prostate Cancer).”
  • Previous studies assessing PCI have demonstrated higher rates of neurocognitive decline with fraction sizes greater than 3 Gy of with concurrent chemotherapy. Previous studies demonstrating a decrease in CNS metastasis following PCI have used doses of 30 to 36 Gy and fraction sizes of 2 to 3 Gy. The recently published Intergroup trial (Le Péchoux C, et al. Lancet Oncol. 2009;10(5):467-74.) assessing PCI in limited stage SCLC demonstrated the superiority of 25 Gy in 2.5 Gy fractions, when compared with 36 Gy in 2 Gy fractions. PCI to a total dose of 25 Gy, in comparison to the 30 Gy used in RTOG 0214, may result in fewer neurocognitive symptoms and an improved quality of life. Additional investigation further determining the optimal dose of PCI for patients with locally advanced NSCLC is warranted but unlikely to occur in the near future due to the poor accrual in this study.
  • At this time, this study is not “practice changing,” as it is unlikely that many practitioners will recommend PCI to their patients with locally advanced NSCLC due to a lack of overall survival benefit demonstrated in this study and persistent concerns regarding the neurocognitive toxicity of PCI. Futures strategies may involve limiting PCI to high-risk patients (non-squamous cell histologies), administering neurologic protector agents (currently being assess in RTOG 0614 with Memantine), using blood-brain barrier penetrating chemotherapy (Temodar), or avoiding prophylactic irradiation of the hippocampus that may be response for the early cognitive decline following PCI.