Inter-patient Variation of Radiosensitivity and Dead-cell Resolving Time in Cervical Cancer

Z. Huang,N. A. Mayr,S. S. Lo,W. T. C. Yuh,G. Jia,C. L. Tosti,S. Sammet,J. Z. Wang

DOI: https://doi.org/10.1016/j.ijrobp.2010.07.968

2010-01-01

Abstract:Tumor response to radiation therapy (RT) is known to vary from patient to patient. In vivo mathematical modeling provides a unique way to derive key radiobiological parameters related to the underlying biological processes that critically influence the tumor's radio-responsiveness, including radiosensitivity and resolution of dead/inactivated cells. The purpose of this study was to investigate the characteristic distributions of tumor radiosensitivity and dead-cell resolving time among different patients with cervical cancer. Quantitative 3D volumetric data of tumor regression were obtained through three sequential MRI studies performed at pre-RT and every 2-2.5 weeks during RT in 110 patients with stage IB2-IVB cervical cancer. A kinetic model incorporating effects of radiation cell killing, tumor repopulation, and resolution of dead/inactivated cells, was applied to the in vivo clinical data to derive the model parameters: the radiosensitivity after 2 Gy (S2) and half time of dead-cell resolving (T1/2). The histograms of these model parameters were generated to determine the characteristic distributions of S2 and T1/2 values from patient to patient. Six-year local tumor control and disease-specific survival data were used as endpoints for outcome prediction (median follow-up 6.2 years). The mathematical model fitted the volume regression data well in each patient, yielding the model parameters S2 and T1/2. The fitting of the distributions of S2 and T1/2 with Gaussian distributions resulted in the mean values and standard deviations of 0.75 and 0.08 for S2, and 9.5 and 6.2 days for T1/2. Large inter-patient variations were found in the two radiobiological parameters, particularly in the T1/2, suggesting to the marked heterogeneity in tumor response to standard radiation therapy. Model parameters correlated with outcome. The 6-year local tumor control rates were 93% vs. 67% for S2<0.77 and S2≥0.77 (p = 0.001), and 88% vs. 30% for T1/2<17 and T1/2≥17 days (p < 0.001), respectively. The 6-year disease-specific survival rates were 79% vs. 54% for S2<0.77 and S2≥0.77 (p = 0.008), and 72% vs. 25% for T1/2<17 and T1/2≥17 days (p < 0.001), respectively. Our preliminary results show wide variation of two major radiobiological parameters, radiosensitivity and dead-cell resolving time, which translate into variations of long-term treatment outcomes in local control and survival in atypical cohort of cervix cancer patients. Such findings demonstrate the potential to model tumor response individually during ongoing radiation therapy and guide the need for more intense therapy in patients at risk of poor outcome of standard care.