Analysis of tumor control probability (TCP) and normal tissue complication probability (NTCP) in 40 prostate cancer patients treated with 2 different techniques of 3D-conformal radiation therapy (3D-CRT)

Abstract

Purpose: To compare 2 regimens of 3D-CRT (3-field vs 6-field technique) in the treatment of prostate cancer, using TCP and NTCP data. Materials & Methods: Between October and December 1997, 20 consecutive patients with clinically localized adenocarcinoma of the prostate (4 T1, 9 T2, 7 T3) were treated with 3D-CRT 3-field technique. For the purpose of this study the records of 20 patients matched for tumor stage (5 T1, 9 T2, 6 T3) and treated with 3D-CRT 6-field technique, were analyzed and served as a basis for comparison. With the 3-field technique, the patients were simulated with rectum and bladder empty in a supine position using custom-made polyurethane foam or thermoplastic casts as immobilization devices. An anterior field (0) and two wedged posterior oblique fields (110 and 250 °) were devised. In order to minimize the dose delivered to the anterior rectal wall and despite a less homogenous dose distribution within the target, the wedges were inserted with the thick end toward the midline. Non-uniform margins were applied around the outline of the target volume, that is, 1 cm towards the rectal wall and 1.5 cm towards the bladder anteriorly and also for the upper and lower portal margins. The shaping of the beam apertures was realized by conventional cerrobend blocks. The total target dose was prescribed at the ICRU point and was 76 Gy for the prostate delivered by a 15 MV linear accelerator in 38 fractions and 56 days (mean). The seminal vesicles were excluded at 70 Gy. The 6-field technique was different only in positioning of the patient (prone) and arrangement of portals (45 °, 90, 135 °, 225*, 270*, 315*). For the calculation of the NTCP values regarding the bladder and rectum we have employed the Lyman's empirical model. The cumulative dose-volume histograms were reduced according to the effective volume model introduced by Kutcher and Burman. Radiobiological data published by Emami was used in this analysis. The TCP calculation was made according to Goitein's model. The mean TCP and NTCP values over the two groups were compared using a pooled two sample t-test. The log percentages were used to ensure that the normality assumptions were valid. Results: TCP values obtained with the two techniques were similar: mean value equals 98.7 for the 3-field technique (range: 97.6-99.4), 98.6 for the 6-field technique (range: 97.5-99.5), meaning a good target coverage with both techniques; p= 0.27, that means no evidence of difference between the two techniques for target coverage. NTCP values were: bladder NTCP 0.96 for 3-field (range: 0.02-4.99) and 0.91 for 6-field (range: 0.02-4.48) (p=0.67), rectum NTCP 7.6 for 3field (range: 2.7-16.9) and 5.7 for 6-field (range: 1.6-11.9) (p=0.051). There is no evidence of difference between the two techniques for bladder complication, but there may be some evidence of difference between the two groups for rectum complications. Conclusions: The two techniques yielded similar results in terms of TCP and bladder NTCP. In terms of rectum NTCP, there was a slight difference in favour of the 6-field technique. Acute toxicity was similar in either groups, but longer follow-up is needed to evaluate the relationship between the mathematical modelling and the rate of late toxicity. The difficulty in extrapolating data from other institutions when calculating TCP and NTCP prompted us to create our own database to evaluate the validity and the predictibility of these mathematical models, considering that different DVH calculation algorithms are usually applied in each center.