Should patient setup in lung cancer be based on the primary tumor? An analysis of tumor coverage and normal tissue dose using repeated positron emission tomography/computed tomography imaging

Abstract

Purpose: Evaluation of the dose distribution for lung cancer patients using a patient setup procedure based on the bony anatomy or the primary tumor.

Methods and materials: For 39 patients with non-small-cell lung cancer, the planning fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) scan was registered to a repeated FDG-PET/CT scan made in the second week of treatment. Two patient setup methods were analyzed based on the bony anatomy or the primary tumor. The original treatment plan was copied to the repeated scan, and target and normal tissue structures were delineated. Dose distributions were analyzed using dose-volume histograms for the primary tumor, lymph nodes, lungs, and spinal cord.

Results: One patient showed decreased dose coverage of the primary tumor caused by progressive disease and required replanning to achieve adequate coverage. For the other patients, the minimum dose to the primary tumor did not significantly deviate from the planned dose: -0.2 ± 1.7% (p = 0.71) and -0.1 ± 1.7% (p = 0.85) for the bony anatomy setup and the primary tumor setup, respectively. For patients (n = 31) with nodal involvement, 10% showed a decrease in minimum dose larger than 5% for the bony anatomy setup and 13% for the primary tumor setup. The mean lung dose exceeded the maximum allowed 20 Gy in 21% of the patients for the bony anatomy setup and in 13% for the primary tumor setup, whereas for the spinal cord this occurred in 10% and 13% of the patients, respectively.

Conclusions: In 10% and 13% of patients with nodal involvement, setup based on bony anatomy or primary tumor, respectively, led to important dose deviations in nodal target volumes. Overdosage of critical structures occurred in 10-20% of the patients. In cases of progressive disease, repeated imaging revealed underdosage of the primary tumor. Development of practical ways for setup procedures based on repeated high-quality imaging of all tumor sites during radiotherapy should therefore be an important research focus.

Figure 1

Example of a patient that had tumor growth in cranial direction. Coverage of this cranial part is not guaranteed by the original treatment plan and cannot be adapted by a bony anatomy or tumor registration based set-up.

Figure 2

Differences in minimum dose expressed as the D99% compared to the planning for the primary tumor (top) and lymph nodes (bottom) for patient set-up based on the bony anatomy or primary tumor. The arrow indicates the large dose difference of the patient which had a growth of the primary tumor.

Figure 3

Individual patient values for the mean lung dose (top) and maximum spinal cord dose (bottom) of the planned dose, bony anatomy and primary tumor based set-up procedure. The planning constraint is shown as a solid line.