Long-Term Outcomes of RTOG 90-03: A Comparison of Hyperfractionation and Two Variants of Accelerated Fractionation to Standard Fractionation Radiotherapy for Head and Neck Squamous Cell Carcinoma

Reviewer: Thomas Dilling, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 18, 2005

Presenter: A. Trotti
Presenter's Affiliation: Moffitt Cancer Center at Univ. South Florida, Tampa, FL
Type of Session: Scientific


  • In the late 1980's, a variety of radiation dose fractionation schedules were in use at different institutions. The standard of care at the time was to treat patients once daily with radiation.  A few  major centers were utilizing twice-daily radiation schemas, in an attempt to improve tumor cell kill.  An open question was whether one scheme was superior to the others, both in terms of local control of tumor and in terms of minimizing long-term radiation side effects in the patient. As a result, the RTOG opened trial 90-03, to look at the major dose fractionation schedules then in use.
  • The initial results, published in 2000 with 1.8 years median follow-up, demonstrated that patients treated with hyperfractionation and accelerated fractionation with concomitant boost had significantly better local-regional control (p = 0.045 and p = 0.050 respectively) than those treated with standard fractionation. There was also a trend toward improved disease-free survival (p = 0.067 and p = 0.054 respectively) although the difference in overall survival was not significant.
  • Reported here are updated results of the trial, with 8.5 years median follow-up.

Materials and Methods
Patients with locally advanced head and neck cancer were randomly assigned to receive radiotherapy delivered with:

  1. standard fractionation (SFX) at 2 Gy/fraction/day, 5 days/week, to 70 Gy/35 fractions/7 weeks;
  2. hyperfractionation (HFX) at 1.2 Gy/fraction, twice daily, 5 days/week to 81.6 Gy/68 fractions/7 weeks;
  3. accelerated fractionation with split (AHFX-S) at 1.6 Gy/fraction, twice daily, 5 days/week, to 67.2 Gy/42 fractions/6 weeks including a 2-week rest after 38.4 Gy; or
  4. accelerated fractionation with concomitant boost (AFX-C) at 1.8 Gy/fraction/day, 5 days/week and 1.5 Gy/fraction/day to a boost field as a second daily treatment for the last 12 treatment days to 72 Gy/42 fractions/6 weeks.
Primary endpoint was loco-regional control (LRC). Secondary endpoints included overall survival (OS), disease-free survival (DFS) and also late-term side effects of radiotherapy in patients. 1113 patients were enrolled in the study, of which 1073 were analyzable. Patients were well-balanced among the treatment arms.


  • LRC for AFHX-S and SFX remained essentially identical, and statistically significantly inferior to AFX-C and HFX.
  • AFX-C and HFX demonstrated a trend toward improved DFS compared to the other two arms, but this did not quite reach statistical significance.
  • There was no difference in OS among the four arms.
  • AFX-C demonstrated a trend toward higher levels of grade 3+ toxicity >4 years out when compared to the other arms.
  • Prevalence of severe side effects seems to diminish overall with time, but this could well represent patient selection biases.
  • Rates of secondary malignancy were high in these patients (approximately 25% at 5 years and 40% at 10 years).
  • There were 11 "late failures" (which the presenter acknowledged might actually represent second malignancies, though this has not been examined closely), of which eight occured >8 years out, with 6 in the HFX arm and 2 in the SFX arm.

Author's Conclusions

  • AFX-C and HFX demonstrate improved LRC of approximately 7%, which was statistically significant.
  • These more aggressive fractionation schedules have previously been shown to have increased complication rates over SFX. With longer follow-up, AFX-C, in particular, shows a trend toward more complications than the other treatment arms.

Clinical/Scientific Implications
Longer-term follow-up of this landmark study is welcome. It is clear that HFX and AFX-C continue to demonstrate improved LRC over SFX or AHFX-S. However, this is at the expense of greater long-term toxicity in these patients. Research has been ongoing to reduce side effects in patients receiving head and neck radiotherapy. These include utilizing more sophisticated techniques of treatment delivery, such as IMRT, as well as ongoing study of  radioprotectants to minimize radiation-induced complications in the patients. However, further work needs to be done in this area.