Biophysical analysis of the acute toxicity of radiotherapy in Hodgkin's lymphoma-a comparison between extended field and involved field radiotherapy based on the data of the German Hodgkin Study Group

Purpose: To determine biophysical parameters from the complication probability data during and after radiotherapy of Hodgkin's lymphoma (HL), based on the number of gastrointestinal side effects that were found in the multicenter HD8 trial of the German Hodgkin Lymphoma Study Group. Methods and Materials: Between 1993 and 1998, 1204 patients with newly diagnosed, histology-proven HL in clinical Stages I/IIA/IIB with defined risk factors and stage IIIA without risk factors were enrolled into the multicenter HD8 study. Patients were randomized to receive two cycles of COPP (cyclophosphamide, vincristine, procarbazine, prednisone) alternating with two cycles of ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine) followed by radiotherapy (RT) of 30 Gy extended field plus 10 Gy to bulky disease (Arm A) or 30 Gy involved field plus 10 Gy to bulky disease (Arm B). For 910 patients, the rates of acute gastrointestinal side effects during and after RT could be determined. Comparison showed differences between Arms A and B (Grade 1-2: 16.6 vs. 3.9; Grade 3-4: 0.9 vs. 0.2; p < 0.001). From the dose-volume histograms for a 'standard patient' (volume intestine 2300 cm3), we determined the normal tissue complication probability (NTCP) (V, D, m, n, TD50 ), the biophysical parameter TD50 , and n (volume dependent) in such a manner that the observed NTCP in Arm A in cases of supradiaphragmatic involvement only and in cases of infradiaphragmatic involvement correlated with the calculated values. Results: Of 1,204 patients randomized, 1,064 patients were informative for the comparison of study arms. The median observation time was 54 months. The overall survival for all eligible patients was 91%, and freedom from treatment failure was 83%. Survival rates at 5 years after start of RT revealed no differences in terms of freedom from treatment failure (85.8% in Arm A, 84.2% in Arm B) and overall survival (90.8% and 92.4%). There were also no differences between the two arms in terms of complete remission, progressive disease, relapse, death, and secondary neoplasias. In contrast, acute side effects, including leukopenia, thrombocytopenia, nausea, gastrointestinal toxicity, and pharyngeal toxicity, were more frequent in the extended field arm. Concerning gastrointestinal toxicity, the different radiation treatment volumes resulted in different NTCPs. On the basis of these findings, values of n = 0.09 and TD50 = 32 Gy were derived. However, this biophysical model is sensitive to variations of the parameters. A deviation of 1% of TD50 results in a deviation of 10% of the NTCP. Conclusion Radiotherapy volume reduction from extended field to involved field after two cycles of COPP/ABVD chemotherapy gives similar results and less toxicity in patients with early-stage, unfavorable HL. Biophysical parameters could be determined from the complication probability data after RT of HL. Because of the exponential dependence, this biophysical model is unstable. It represents a 'start model' until further data can be incorporated