Randomized Phase III Comparison of Standard Dose (60Gy) Versus High-Dose (74Gy) Conformal Chemoradiotherapy +/- Cetuximab for Stage IIIA/B Non-Small Cell Lung Cancer: Preliminary Findings on Radiation Dose in RTOG 0617

Reporter: Surbhi Grover, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 3, 2011

Presenter: Jeffrey Bradley, MD


  • Dose escalation has been a controversial topic in definitive treatment of non-small cell lung cancer (NSCLC) with previous studies showing only 20% pathological complete response with 60 Gy and concurrent chemotherapy.
  • Randomized trials of concurrent chemoradiation, although better than sequential chemotherapy and radiation, still show poor overall survival (OS) at 5 years. Recently published 94-10 showed 16% survival at 5 years with concurrent chemotherapy and radiation to 60 Gy with median survival of 17 months.
  • There is phase II data (CALGB 30105) that suggests improved survival, median survival 24 months, with concurrent carboplatin/taxol with dose escalation to 74Gy.
  • RTOG 0617 was designed to study the role of: 1) dose escalation 2) cetuximab, EGFR monoclonal antibody in improving OS of stage III NSCLC patients.
  • This presentation described the results of the dose escalation question. Results of the cetuximab arms are still pending.

Methods and Statistics

  • Patients with newly diagnosed stage III NSCLC, with no supraclavicular or contralateral hilar disease, ECOG 0-1 PS and decent PFTs (FEV1>1.2 L/sec or >50% of predicted) were randomized.
  • Stratification was done by radiation techinque (3DCRT vs IMRT), performance status, histology and whether or not PET staging was performed.

This is a 2x2 factorial, 4 arm study. Arms are as follows:

  1. Carbo/taxol with 60 Gy concurrent radiation
  2. Carbo/taxol concurrent with 60 Gy + consolidation carbo/taxol and cetuximab
  3. Carbo/taxol with 74 Gy concurrent radiation
  4. Carbo/taxol + cetuximab concurrently with 74 Gy followed by consolidation chemotherapy

Accrual began in November 2007 with target accrual of 500 patients. With the primary endpoint of OS, hypothesis was to see a median survival improvement from 17 months to 24 months (with both dose escalation and cetuximab). Study was designed to detect 0.0125 significance (one-sided logrank test) with 80% power at 399 events. Interim analysis was planned at 85, 170 and 255 events. This data is from the first interim analysis at 85 events.


From November 2007-June 2011 464 patients were accrued. Of these, 426 patients were eligible.

  • Pretreatment characteristics for patients in both arms, 60 Gy vs 74 Gy, were balanced including in regards to RT technique, number of patients who had PET staging, histology, race and gender. About 45% of patients in both arms had IMRT, and 90% had PET staging.
  • As for the dose, both the arms were balanced. Mean lung V20 Gy was about 30% in both arms. Mean median esophageal dose was about 28% in both arms. Median CTV--> PTV margin was 7mm (minimum of 0.5cm-1.5cm was required).
  • In regards to toxicity, there was no significant difference in grade 3, 4 and 5 toxicity between the two arms. As for grade 5 events, there were four events in 60 Gy and eight events in 74 Gy arm. There were two radiation pneumonitis deaths in each arm.
  • Median follow up is 11.7 months in the 60 Gy arm vs 10.3 months in 74 Gy arm.
  • Survival: OS at 12 months is 81% in the 60 Gy arm vs 74% in the 74 Gy arm (p=0.02). Median survival was 21.7 months vs 20.7 months in the 60Gy vs 74Gy arms respectively.
  • On univariate analysis, smaller GTV, total lung volume, age, non-squamous histology and gender were significant in predicting over all survival.
  • On multivariate analysis, standard dose, non-squamous histology and smaller GTV were significant predictors of OS.

Authors conclusions

  • High dose radiation arm was closed early for futility; there was no benefit to dose escalation. Trial is still open at 60 Gy arm to answer the cetuximab question.
  • Toxicity was similar between the two arms.
  • Factors predicting survival were standard radiation dose, smaller GTV and non-squamous histology.
  • At this point there are no factors detected that could predict the causes for futility of the dose escalation.

Clinical and Scientific Implications

There is a no level I evidence for dose escalation for definitive treatment of NSCLC. It is important to note that both arms of this study, at one year had higher OS (OS at 12 months: 81% for 60 Gy arm and 70% for 74 Gy arm) than the concurrent chemoRT arm of RTOG 94-10 (OS at 12 months 62%), which may be the result of improved RT techniques and staging modalities.

It is interesting to note that the high dose arm actually did worse than the standard dose arm. What may be causing this?

  1. Even though toxicity was not statistically significant in the two groups, rates of grade 5 toxicity were 2% higher in the high dose arm. Enough to explain the 10% survival difference between the two arms? Most likely not.
  2. On multivariate analysis, non-squamous histology and smaller GTV size were significant predictors of survival. Although not statistically significant, there were more patients with squamous histology in the high dose arm compared to standard dose arm. As for GTV, there were actually more patients with smaller tumors in the high dose arm, so unlikely a factor at play in explaining outcomes of the high dose arm.
  3. This study allows up to <10% weight loss and there is no data on smoking status, other co-morbidities or quality of life of the patients (a significant predictor of survival), which could be imbalanced in the two arms and possibly confounding the results.
  4. As for technical aspects, study allowed use of higher energy beams adding to the unpredictability of dose distribution. Minimum dose was allowed to be 93% and maximum dose was allowed to be 125%, which may lead to a large dose variation within the PTV. Median CTV--> PTV margin was about 7mm, which may not be sufficient. Potentially, variation in the technical issues between the two arms could have confounded the results.

More data on patterns of failure will be important in guiding our interpretation of these results. If local control is better in the high dose arm, maybe there is still a role for dose escalation with improved systemic agents. If the local control is worse in the high dose arm that would suggest that improvement in technical issues discussed above are even more essential.

In conclusion, as of now, there is no evidence for dose escalation, but we have a long way to go in terms of improving techniques and exploring novel agents that may improve survival in NSCLC.

Frequently Asked Questions