Dexamethasone and supportive care with or without whole brain radiotherapy in treating patients with non-small cell lung cancer with brain metastases unsuitable for resection or stereotactic radiotherapy (QUARTZ): results from a phase 3, non-inferiority, randomised trial

Abstract

Background: Whole brain radiotherapy (WBRT) and dexamethasone are widely used to treat brain metastases from non-small cell lung cancer (NSCLC), although there have been no randomised clinical trials showing that WBRT improves either quality of life or overall survival. Even after treatment with WBRT, the prognosis of this patient group is poor. We aimed to establish whether WBRT could be omitted without a significant effect on survival or quality of life.

Methods: The Quality of Life after Treatment for Brain Metastases (QUARTZ) study is a non-inferiority, phase 3 randomised trial done at 69 UK and three Australian centres. NSCLC patients with brain metastases unsuitable for surgical resection or stereotactic radiotherapy were randomly assigned (1:1) to optimal supportive care (OSC) including dexamethasone plus WBRT (20 Gy in five daily fractions) or OSC alone (including dexamethasone). The dose of dexamethasone was determined by the patients' symptoms and titrated downwards if symptoms improved. Allocation to treatment group was done by a phone call from the hospital to the Medical Research Council Clinical Trials Unit at University College London using a minimisation programme with a random element and stratification by centre, Karnofsky Performance Status (KPS), gender, status of brain metastases, and the status of primary lung cancer. The primary outcome measure was quality-adjusted life-years (QALYs). QALYs were generated from overall survival and patients' weekly completion of the EQ-5D questionnaire. Treatment with OSC alone was considered non-inferior if it was no more than 7 QALY days worse than treatment with WBRT plus OSC, which required 534 patients (80% power, 5% [one-sided] significance level). Analysis was done by intention to treat for all randomly assigned patients. The trial is registered with ISRCTN, number ISRCTN3826061.

Findings: Between March 2, 2007, and Aug 29, 2014, 538 patients were recruited from 69 UK and three Australian centres, and were randomly assigned to receive either OSC plus WBRT (269) or OSC alone (269). Baseline characteristics were balanced between groups, and the median age of participants was 66 years (range 38-85). Significantly more episodes of drowsiness, hair loss, nausea, and dry or itchy scalp were reported while patients were receiving WBRT, although there was no evidence of a difference in the rate of serious adverse events between the two groups. There was no evidence of a difference in overall survival (hazard ratio 1·06, 95% CI 0·90-1·26), overall quality of life, or dexamethasone use between the two groups. The difference between the mean QALYs was 4·7 days (46·4 QALY days for the OSC plus WBRT group vs 41·7 QALY days for the OSC group), with two-sided 90% CI of -12·7 to 3·3.

Interpretation: Although the primary outcome measure result includes the prespecified non-inferiority margin, the combination of the small difference in QALYs and the absence of a difference in survival and quality of life between the two groups suggests that WBRT provides little additional clinically significant benefit for this patient group.

Funding: Cancer Research UK, Medical Research Council Clinical Trials Unit at University College London, and the National Health and Medical Research Council in Australia.

Figure 1

Trial profile WBRT=whole brain radiotherapy. OSC=optimal supportive care.

Figure 2

Components of quality-adjusted life-years (QALY) (A) Overall survival from randomisation. (B) Quality of life. The average utility score is calculated from all surviving and uncensored patients for every time at which any patient completed the EQ-5D questionnaire. If a patient has not completed the questionnaire on a particular day, their score is imputed by assuming a straight line connecting their closest utility scores before and after the day in question. (C) Quality-adjusted life-years. The survivor function is multiplied by the average utility score at each time that either the survivor function or the average utility score changes. The area under the resultant step function is the mean QALY for each treatment group. Graphs are only displayed up to 56 weeks for presentation purposes and due to the small number of patients beyond this point.

Figure 3

Forest plot of overall survival by patient characteristics All hazard ratios are obtained from Cox proportional hazard models with adjustment for randomised group only. KPS=Karnofsky Performance Status. NSCLC=non-small cell lung cancer. RPA=recursive partitioning analysis. GPA=graded prognostic assessment. WBRT=whole brain radiotherapy. OSC=optimal supportive care.