Consistency in reference radiotherapy dosimetry: resolution of an apparent conundrum when (60)Co is the reference...

Substantial changes in ion chamber perturbation correction factors in (60)Co γ-rays, suggested by recent Monte Carlo (MC) calculations, would cause a decrease of about 1.5% in the reference dosimetry of all types of charged particles (electrons, protons and heavier ions) based on calculated kQ values. It has gone largely unnoticed that the ratio of calibration coefficients ND, w, Co60 and NK, air, Co60 yields an experimental value of Fch, Co60 = (sw-air pch)Co60 through ND, air, Co60. Coefficients provided by the IAEA and traceable to the BIPM for 91 NE-2571 chambers result in an average Fch, Co60 which is compared with published (and new) MC simulations and with the value in IAEA TRS-398. It is shown that TRS-398 agrees within 0.12% with the experimental Fch, Co60. The 1.5% difference resulting from MC calculations (1.1% for the new simulations) cannot be justified using current fundamental data and BIPM standards if consistency in the entire dosimetry chain is sought. For photons, MC kQ factors are compared with TRS-398. Using the same uncertainty for Wair, the two sets of data overlap considerably. Experimental kQ values from standards laboratories lie between the two sets of calculated values, showing no preference for one set over the other. Observed chamber-to-chamber differences, that include the effect of waterproof sleeves (also seen for (60)Co), justify the recommendation in TRS-398 for kQ values specifically measured for the user chamber. Current developments on I-values for the stopping powers of water and graphite are presented. A weighted average Iwater = 78 ± 2 eV is obtained from published experimental and DRF-based values; this would decrease sw-air for all types of radiotherapy beams between 0.3% and 0.6%, and would consequently decrease the MC derived Fch, Co60. The implications of a recent proposal for Igraphite = 81 eV are analysed, resulting in a potential decrease of 0.7% in NK, air, Co60 which would raise the experimental Fch, Co60; this would result in an increase of about 0.8% in the current TRS-398 value when referred to the BIPM standards. MC derived Fch, Co60 using new stopping powers would then agree at a level of 0.1% with the experimental value, confirming the need for consistency in the dosimetry chain data. Should world average standards be used as reference, the figures would become +0.4% for TRS-398 and -0.3% for the MC calculation. Fch, Q calculated for megavoltage photons using new stopping powers would decrease by between 0.2% and 0.5%. When they enter as a ratios in kQ, differences with MC values based on current key data would be within 0.2% but their discrepancy with kQ experimental photon values remains unresolved. For protons the new data would require an increase in Wair, Q of about 0.6%, as this is inferred from a combination of calorimetry and ionometry. This consistent scenario would leave unchanged the current TRS-398 kQ (NE-2571) data for protons, as well as for ions heavier than protons unless new independent Wair, Q values become available. Also in these advanced radiotherapy modalities, the need for maintaining data consistency in an analysis that unavoidably must include the complete dosimetry chain is demonstrated.