Randomized Trials and IMRT: Overkill or Overdue?

S. Jack Wei, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 22, 2003

Moderator: Theodore L. Phillips, MD., University of California, San Francisco

Soeren M. Bentzen, PhD., DSc., Gray Laboratory Cancer Research Trust, Mount Vernon Hospital, Northwood, Middlesex, United Kingdom

IMRT is an advanced technology for radiation therapy (RT), using non-uniform radiation beam intensities incident upon the patient.
From a dosimetric and operational point of view, IMRT is superior to forward-planned treatment, including 3DCRT.
Despite, the dosimetric advantages of IMRT, the question remains: Will IMRT improve clinical outcomes?
Various schemes for levels of clinical evidence support of a treatment have been proposed; most define level I evidence as large, adequately powered randomized trials or a metaanalysis supporting the use of treatment. Level IV evidence can be defined as single-arm non-randomized trials comparing results to historical controls.
Clinical outcome data for IMRT is almost solely level IV evidence.
Level IV evidence does not equal Level I evidence.
As a historical example, the use of continuous hyperfractionated accelerated radiation therapy (CHART) for head and neck (H&N) cancers was studied and compared to historical controls treated with conventionally fractionated RT matched for tumor site, T stage, and N stage. These comparisons showed improved outcome with CHART; however, a randomized trial of CHART showed no difference compared to conventional fractionation.
The above example demonstrates the importance of randomized trials.
In conclusion, evidence-based IMRT is not showing improved distribution but a therapeutic gain; it is not showing decreased toxicity, but doing so without compromising local control, it is not increasing local control, but doing so without increasing toxicity.