/     /     /  

Can We Predict Radiation (RT)-Induced Changes in Pulmonary Function Tests (PFTs) Based on the Lung Dose-Volume Histogram?

Theodore Robnett, MD
OncoLink Assistant Editor
Ultima Vez Modificado: 1 de noviembre del 1999

Presenter: Ming Fan, M.D.
Affiliation: Duke University Medical Center

Summary:
Radiation induced pulmonary changes can cause serious, even life-threatening problems in patients whose lung function is already compromised by tumor burden and smoking. There are many studies which qualitatively document pulmonary toxicity relative to dosing parameters. The present study attempts to predict quantitative changes in PFTs based on these parameters.

Methods:
Ninety-six of 185 patients (68% with lung cancer) were prospectively evaluated with PFTs prior to and 6 months following RT. Thirty-four patients received chemotherapy. Criteria for selection included no evidence of intra-thoracic recurrence at follow up. Predicted PFTs were based on the summation of products derived from dose volume histograms (DVHs) and reduced lung volumes. An alternate formula used perfused lung volumes (DFH) instead of total lung volumes. Predicted and measured changes were compared using linear regression. A multivariate analysis on clinical variables including age, tumor site and baseline PFTs was also performed.

Results:

• Although there was a statistically significant correlation between predicted and actual outcomes (p = 0.005), maximum decline in PFT measurements were not well predicted by DVH or DFH (R2 = 0.3). This may be partly explained by the observation that many patients showed improved PFTs after radiation therapy.
• The presenter stated that a multivariate model provided better predictions, but unfortunately he did not provide further details regarding this.

Clinical/Scientific Implications:

• Radiation related pulmonary injury has been previously related to humoral and clinical factors.
• Although dose and distribution to tissue play a role in this important toxicity, the present study would suggest that multiple factors influence PFT changes and ultimate patient outcome.
• The presenter suggests that an incorporation of tumor location into modeling may improve the ability to predict PFT outcome in patients receiving thoracic radiation therapy.

ASTRO coverage is sponsored, in part, by Varian Medical Systems, Inc.