Dosimetric effect of respiratory motion in external beam radiotherapy of the lung

James Mechalakos, Ellen Yorke, Gikas S. Mageras, Agung Hertanto, Andrew Jackson, Ceferino Obcemea, Kenneth Rosenzweig, C. Clifton Ling

Research output: Contribution to journalArticlepeer-review

47 Scopus citations


Background and purpose To study the effect of breathing motion on gross tumor volume (GTV) coverage for lung tumors using dose-volume histograms and relevant dosimetric indices. Patients and methods Treatment plans were chosen for 12 patients treated at our institution for lung carcinoma. GTV volumes of these patients ranged from 1.2 to 97.3 cm3. A margin of 1-2 cm was used to generate the planning target volume (PTV). Additional margins of 0.6-1.0 cm were added to the PTV when designing treatment portals. For the purposes of TCP calculation, the prescription dose was assumed to be 70Gy to remove the effects of prescription differences. Setup error was incorporated into the evaluation of treatment plans with a systematic component of σRL=0.2cm, σAP=0.2cm, and σ SI=0.3cm and a random component of σRL=0.3cm, σAP=0.3cm, and σSI=0.3cm. Breathing motion was incorporated into these plans based on an independent analysis of fluoroscopic movies of the diaphragm for 7 patients. The systematic component of breathing motion (σRL=0.3cm, σAP=0.2cm, and σSI=0.6cm) was incorporated into the treatment plans on a slice by slice basis. The intrafractional component of breathing motion (σRL=0.3cm, σAP=0.2cm, and σSI=0.6cm) was incorporated by averaging the dose calculation over all displacements of the breathing cycle. Each patient was simulated 500 times to discern the range of possible outcomes. The simulations were repeated for a worst case scenario which used only breathing data with a large diaphragmatic excursion, both with and without intrafractional breathing motion. Results Dose to 95% of the GTV (D95), volume of the GTV receiving 95% of the prescription dose (V95), and TCP changed an average of -1.4±4.2, -1.0±3.3, and -1.4±3.8%, respectively, with the incorporation of normal breathing effects. In the worst case scenario (heavy breathers), D95 and V95 changed an average of -9.8±10.1 and -8.3±11.3%, respectively, and TCP changed by -8.1±9.1%. GTVs with volumes greater than 60 cm 3 showed stronger sensitivity to breathing especially if the shape was non-ellipsoidal. In the normal breathing case, the probability of a decrease in D95, V95, or TCP of a magnitude greater than 10% is less than 4%, and in the worse case scenario this probability is approximately 30-40% with intrafractional breathing motion included, and less than 10% with intrafractional breathing motion not included. Conclusions With the PTV margins routinely used at our center, the effects of normal breathing on coverage are small on the average, with a less than 4% chance of a 10% or greater decrease in D95, V95, or TCP. However, in patients with large respiration-induced motion, the effect can be significant and efforts to identify such patients are important.

Original languageEnglish
Pages (from-to)191-200
Number of pages10
JournalRadiotherapy and Oncology
Issue number2
StatePublished - May 2004
Externally publishedYes


  • Breathing motion
  • DVH
  • Lung radiotherapy
  • Monte Carlo
  • TCP


Dive into the research topics of 'Dosimetric effect of respiratory motion in external beam radiotherapy of the lung'. Together they form a unique fingerprint.

Cite this