Purpose: Partial coirradiation of both kidneys is an unavoidable consequence of adequate dose delivery in radiation therapy of para-aortic lymph nodes (PLN). Depending on total dose anteroposterior/posteroanterior (AP/PA), opposed-fields or multifield techniques are used. To optimize the treatment of potentially tumor-affected PLN with minimal kidney involvement, we calculated normal tissue complication probabilities (NTCPs) of coirradiated kidneys for four common irradiation techniques used in the PLN area.
Methods and materials: Planning target volume (PTV) delineation was performed in computed tomography scans of 21 patients with a lateral safety margin of 3 cm from the aorta and 2 cm aside the vena cava. Ventral and dorsal margins of the PTV were delineated 2 cm from the vessels. As previously shown (Nevinny-Stickel M, et al. Int J Radiat Oncol Biol Phys 2000;48:147-151), PTVs optimized by these altered delineations permit inclusion of at least 97% of potentially involved PLN in contrast to standard delineations based on bony structures that are more likely to miss affected lymph nodes. The present study compared NTCPs for individual PTV-based treatment planning with NTCPs for standard planning based on bony structures. For each patient, four hypothetical treatment plans were created: (A) standard AP/PA opposed fields technique with lateral field margins along the tips of the transverse processes of the vertebral bodies; (B) individually planned AP/PA opposed fields with lateral field margins according to the optimized PTV; (C) standard four-field box technique with lateral width as described for (A), with dorsal borders at the center of the vertebral bodies and ventral margins 3 cm in front of the vertebrae; and (D) individually planned four-field box with lateral field margins according to the optimized PTV. Calculation of irradiation-induced complication probability values for nonuniform kidney irradiation was performed for model doses 19.8 Gy, 30.6 Gy, and 50.4 Gy according to the Lyman-Wolbarst model.
Results: No dose showed a statistically significant difference (p < 0.00833, corrected for six multiple interrelated comparisons) in the median of total organ kidney NTCPs between techniques A, C, and D, with technique D intermediately ranging between technique A and C (e.g., for 50.4 Gy: A: median, 0.39; range, 0.01-0.83; C: median, 0.27 range; 0.05-0.68; D: 0.36; range, 0.03-0.72). In comparison to techniques A, C, and D, the individually planned AP/PA opposed-fields technique (B) was accompanied by significantly higher and intolerable overall kidney NTCP rates (e.g., for 50.4 Gy: median, 0.68; range, 0.01-0.99).
Conclusion: Conformal four-field planning with individually optimized PTVs (D) resulted in only moderate tissue complication probabilities in both kidneys with the advantage of providing significantly greater inclusion of potentially involved PLNs in comparison to accepted standard procedures (A and C).