3730 - Is the a/ß of Cervical Cancer Low? New Clinical Estimates of the a/ß of Cervical Cancer

Purpose/Objective(s):

The a/ß of cervical cancer is currently assumed to be 10 Gy with little to no evidence supporting such assumption. To date, no clinical estimates of a/ß for cervical cancer have been derived under the current standard of treatment of external beam therapy combined with high dose rate (HDR) brachytherapy. In addition, accounting for clinical considerations such as inter-patient heterogeneity of radiosensitivity and adjuvant treatments should give more reasonable estimates of a/ß. With these considerations the a/ß of cervical cancer is expected to be much lower than the assumed 10 Gy.

Materials/Methods:

A retrospective analysis of 180 cervical cancer patients was conducted, with local control rates for follow-up periods of one to five years. The Poisson TCP model was then fit to the outcome data. Two versions of the Poisson model were assessed: First a non-heterogenous Poisson model based on the linear quadratic (LQ) model and then a version that incorporated inter-patient heterogeneity of the parameters a and ß. The heterogenous model included variables to parameterize the population variance of those parameters. As 54 patients received radical hysterectomies, both Poisson models were also fit to the non-hysterectomy patients to assess its influence. A logistic TCP model was also fitted to the data, with one version only considering dose and treatment time, and another model incorporating tumor stage and chemotherapy and surgery. All models were based on the LQ model with inclusion of cellular repopulation.

Results:

The non-heterogenous and heterogeneous Poisson models, when fit to the entire patient dataset gave a/ß estimates of 87.0 Gy and 7.96 Gy respectively. Fitted to just the non-surgery cohort, the non-heterogenous and heterogenous Poisson models gave values of 23.75 Gy and 3.53 Gy respectively. The first logistic TCP model gave an a/ß value of 27.96 Gy. The logistic model including tumor stage and adjuvant treatments gave a much lower estimate of 2.62 Gy. When accounting for the increased local control caused by surgical hysterectomies, both Poisson and logistic methods gave very similar estimates, averaging to about 3.08 Gy.

Conclusion:

These are novel clinical estimates of a/ß, the first to analyze patients treated under the standard of external beam plus HDR brachytherapy, and the first clinical study to estimate a low a/ß for cervical cancer. When accounting inter-patient heterogeneity and the influence of adjuvant treatments, predictions averaged to a value of 3.08 Gy, well below the assumed 10 Gy. These new low a/ß estimates may have significant implications for the future of cervical cancer treatment, and the effect of inter-patient heterogeneity and adjuvant treatments must be seriously considered if we wish to derive more accurate clinical estimates of a/ß.