Prediction of radiotherapy response using SF2: is it methodology or mythology?

The reproducibility and relationship between single and fractionated dose estimates of the surviving fraction at 2 Gy

PURPOSE

To evaluate the reproducibility and relationship between the surviving fraction at 2 Gy obtained from single-graded dose and multifraction survival curve assays.

METHODS AND MATERIALS

Single-graded dose and multifraction survival curve assays were concurrently performed in five human cell lines. For the multifraction studies, five to six doses at 2 Gy per fraction were administered with a time interval of 5.5 h between fractions. All surviving fraction data was corrected for multiplicity and cell proliferation during treatment. Replicate experiments were performed for each cell line.

RESULTS

The precision of the Sf2 estimates obtained from multifraction studies was approximately three times greater than was obtained in the single-dose studies. For the single-dose studies, the average difference between the Sf2s obtained in the initial vs. repeat experiment was 0.11; for the fractionated-dose studies the difference was significantly reduced to 0.032. A rank-order correlation was not obtained between the 2 Gy Sfs in the initial and repeat single-dose assays; in the multifraction studies, the correlation was significant (p < 0.05). The rank-order correlation between the single- and fractionated-dose Sf2 assays was significant only when the two sets of assays were pooled; however, the coefficient of correlation remained low (R(2) = 0.50).

CONCLUSIONS

The precision of the estimates of the surviving fraction at 2 Gy, obtained from multifraction assays, was substantially greater than was obtained in single dose-survival curve assays.