Table 6Overall grade of strength of evidence, IMRT vs. 2DRT

Key QuestionStudy DesignRisk of BiasConsistencyDirectnessPrecisionOverall Grade/Conclusion
1. What is the comparative effectiveness of IMRT and 2DRT regarding quality of life and adverse events?Among 22 comparative studies addressing IMRT and 2DRT, 2 were RCTs, and 20 were observational, of which 5 were prospective designs.A high risk of bias was observed throughout this set of studies. One RCT was rated as fair, while all other studies were rated as poor by the USPSTF framework.

Risk of bias was due to:
  • uncertainty re: comparable groups
  • uncertainty re: blinding of outcome assessors noncontemp- oraneous treatment groups or unclear,
  • lack of well-done multivariable analyses to adjust for confounding among observational studies and
  • no ITT analysis in one RCT
Consistent results were observed for two outcomes:
  • quality of life (3 studies); and
  • late xerostomia (8 of 9 studies), particularly domains most related to xerostomia
Statistically significant or otherwise moderate to large differences favored IMRT.

Although the observational studies are not well designed to control for bias and confounding, it is unlikely that there was systematic imbalance of patients with a lower susceptibility to xerostomia in the IMRT groups. Susceptibility to xerostomia is common in the head/neck cancer population due to cancer site and prior and concurrent treatments, and sometimes due to older age and chronic medications

Inconsistent results were observed for these outcomes:
  • acute xerostomia;
  • acute mucositis;
  • late mucositis;
  • acute dysphagia;
  • late dysphagia
  • acute skin toxicity;
  • late skin toxicity; and
  • late osteoradionecrosis and bone toxicity.
Some of the strongest results were also found in studies with substantial methodological weaknesses.

Of six comparative studies reporting patient survival, one reported a statistically significant result; the difference was large and favored IMRT.

Of the five comparative studies reporting tumor control, none reported statistically significant differences between IMRT and 2DRT.
Although direct evidence was available for all outcomes considered under this Key Question, complementary evidence from the comparison between IMRT and 3DCRT provides strong but indirect support for conclusions.Confidence intervals around observed treatment effects were not reported.

Although we could not arrive at a pooled estimate of the effect or a confidence interval for the effect, the consistent direction and moderate-to-large differences.
The strength of the body of evidence is moderate for IMRT reducing late xerostomia and improving quality of life domains related to xerostomia compared with 2DRT. The direct evidence reviewed on IMRT vs. 2DRT, although of limited quality, suggests a true effect in favor of IMRT. Indirect evidence, inference from comparison of IMRT vs. 3DCRT, provides additional support for this conclusion. The advantage of IMRT over 3DCRT is due to greater conformality of radiation delivery. Similarly, IMRT is inferred to be superior to 2DRT because 2DRT is less conformal than 3DCRT.

The strength of evidence is insufficient to draw conclusions about the comparative impact of IMRT and 2DRT for other adverse events.
2. What is the comparative effectiveness of IMRT and 2DRT regarding tumor control and patient survival?Key Question 1 and Key Question 2 were addressed by a common set of studies.As these are the same studies considered for Key Question 1, the risk of bias is high, as noted above.

Except for one fair quality randomized trial, all studies were rated as poor quality by the USPSTF framework and therefore had an inherent risk of bias.

Moreover, estimating between-group differences in disease-specific and overall survival is more complex than for adverse events. Observational studies can only be informative if there is detailed information about long-term losses to followup and well-done multivariable adjustment for confounding.
The evidence does not show consistently significant between-group differences for patient survival and tumor control.

Of six comparative studies reporting patient survival, one reported a statistically significant result; the difference was large and favored IMRT.

Of five comparative studies reporting tumor control, none reported statistically significant differences between IMRT and 2DRT.
Direct evidence is available for overall survival.

Tumor control measures are intermediate outcomes, and are informative to the extent that they predict differences in disease-specific or overall survival.
Confidence intervals around observed treatment effects were not reported.

Direction of effect cannot be determined.
The strength of the body of evidence is insufficient for tumor control and patient survival..

No conclusions on tumor control or survival can be drawn from the body of evidence comparing IMRT versus 2DRT.
3. Patient and tumor characteristics affecting outcomesNo comparative studies addressed this Key Question.NANANANAThe strength of evidence is insufficient, thus no conclusions can be reached.
4. Radiotherapy or physician characteristics affecting outcomesNo comparative studies addressed this Key Question.NANANANAThe strength of evidence is insufficient, thus no conclusions can be reached.

Abbreviations: ITT: intention to treat ; NA: not applicable; RCT: randomized, controlled trial

From: Results

Cover of Comparative Effectiveness and Safety of Radiotherapy Treatments for Head and Neck Cancer
Comparative Effectiveness and Safety of Radiotherapy Treatments for Head and Neck Cancer [Internet].
Comparative Effectiveness Reviews, No. 20.
Samson DJ, Ratko TA, Rothenberg BM, et al.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.