PURPOSE: To express the magnitude of the contribution of chemotherapy to local tumor control in chemoradiotherapy cervical cancer trials in terms of the concept of the biologically effective dose. METHODS AND MATERIALS: The local control rates of both arms of each study (radiotherapy vs. radiotherapy plus chemotherapy) reported from randomized controlled trials of concurrent chemoradiotherapy for cervical cancer were reviewed and expressed using the Poisson model for tumor control probability (TCP) as TCP = exp(-exp E), where E is the logarithm of cell kill. By combining the two TCP values from each study, we calculated the chemotherapy-related log cell kill as Ec = ln[(lnTCP(Radiotherapy))/(lnTCP(Chemoradiotherapy))]. Assuming a range of radiosensitivities (alpha = 0.1-0.5 Gy(-1)) and taking the calculated log cell kill, we calculated the chemotherapy-BED, and using the linear quadratic model, the number of 2-Gy fractions corresponding to each BED. The effect of a range of tumor volumes and radiosensitivities (alpha Gy(-1)) on the TCP was also explored. RESULTS: The chemotherapy-equivalent number of 2-Gy fractions range was 0.2-4 and was greater in tumors with lower radiosensitivity. In those tumors with intermediate radiosensitivity (alpha = 0.3 Gy(-1)), the equivalent number of 2-Gy fractions was 0.6-1.3, corresponding to 120-260 cGy of extra dose. The opportunities for clinically detectable improvement are only available in tumors with intermediate radiosensitivity with alpha = 0.22-0.28 Gy(-1). The dependence of TCP on the tumor volume decreases as the radiosensitivity increases. CONCLUSION: The results of our study have shown that the contribution of chemotherapy to the TCP in cervical cancer is expected to be clinically detectable in larger and less-radiosensitive tumors.
PURPOSE: To express the magnitude of the contribution of chemotherapy to local tumor control in chemoradiotherapy cervical cancer trials in terms of the concept of the biologically effective dose. METHODS AND MATERIALS: The local control rates of both arms of each study (radiotherapy vs. radiotherapy plus chemotherapy) reported from randomized controlled trials of concurrent chemoradiotherapy for cervical cancer were reviewed and expressed using the Poisson model for tumor control probability (TCP) as TCP = exp(-exp E), where E is the logarithm of cell kill. By combining the two TCP values from each study, we calculated the chemotherapy-related log cell kill as Ec = ln[(lnTCP(Radiotherapy))/(lnTCP(Chemoradiotherapy))]. Assuming a range of radiosensitivities (alpha = 0.1-0.5 Gy(-1)) and taking the calculated log cell kill, we calculated the chemotherapy-BED, and using the linear quadratic model, the number of 2-Gy fractions corresponding to each BED. The effect of a range of tumor volumes and radiosensitivities (alpha Gy(-1)) on the TCP was also explored. RESULTS: The chemotherapy-equivalent number of 2-Gy fractions range was 0.2-4 and was greater in tumors with lower radiosensitivity. In those tumors with intermediate radiosensitivity (alpha = 0.3 Gy(-1)), the equivalent number of 2-Gy fractions was 0.6-1.3, corresponding to 120-260 cGy of extra dose. The opportunities for clinically detectable improvement are only available in tumors with intermediate radiosensitivity with alpha = 0.22-0.28 Gy(-1). The dependence of TCP on the tumor volume decreases as the radiosensitivity increases. CONCLUSION: The results of our study have shown that the contribution of chemotherapy to the TCP in cervical cancer is expected to be clinically detectable in larger and less-radiosensitive tumors.