How to use the Paris system in the year 2001?

Abstract

In this issue of Radiotherapy& Oncology, Pieters et al. [4] compared the dosimetry of four different implants for breast cancer. The ®rst implant was made according to the Paris system rules, and the fourth one was done with today's improvements in brachytherapy (stepping sources, optimisation, volume calculation, reference isodose adapted to the planned target volume). The comparison between the implants was made according to the actual concepts of ICRU, using the planned target volume (PTV) as a reference. The fourth implant offered the same quality of target coverage, but decreases signi®cantly the volume of normal breast tissue irradiated to a high dose, so decreases the probability of late ®brosis [1]. The main usefulness of the Paris system [2] was to allow to foresee what will be the dose distribution around multiple sources, and to have some rough guidelines to implant. Forty years after its conception and its large acceptance in the brachytherapist community, what remains of the Paris system and how to use it in the era of stepping sources? To answer this question, the best way seems to review one by one the rules of the Paris system: First rule: the radioactive lines must be parallel and equidistant: for most of anatomical localisation treated with intersticial radioactive wires, this rule of implantation is still actually the easiest way to obtain a good implant (that is an implant which cover fully the target volume without irradiating a large volume of normal tissue) while minimizing the size of hot spots. It allows the prediction of what will be the dose distribution around multiple sources, whatever the dosimetric system. The intersource spacing is an important parameter to consider in this type of implant [5]. In the future, when all the implants will be done under CT-scan visualization and with real-time dosimetric systems, maybe some other geometrical sets could be used. Second rule: the activity must be uniform all along the lines. The use of stepping sources without the same dwell times at each position does not ful®l this rule. But, modifying the dwell times allow to perform optimisation. The virtues of optimization, particularly the geometric optimisation, is actually well demonstrated in the paper of Pieters et al. [4] and others [7,8]: it allows the disappearance of the invagination of the reference isodose in the plane parallel to the needles, and so the decrease of the active lengths. So, this second rule is actually no longer valid. However, it must be kept in mind that optimization cannot correct all the geometric errors of an implant: particularly, a geographic miss, that is a part of the target volume outside the volume de®ned by the needles, is often impossible to ®x, even with the best dosimetric software. It is impossible to design a good distribution of dose with a poor geometry of sources. We can only improve slightly imperfect implant. Third rule: the center of the active lines must be placed in a same plane called the central plane. In this central plane, the mean central dose is calculated, and the reference isodose (85% of the mean central dose) determined. In the work of Pieters et al. [4], as in many institutions, the use of a transversal central plane to de®ne the mean central dose and the reference isodose, remains the rule. However, with modern dosimetric software, many other planes are available, particularly the sagittal plane; it is conceivable to de®ne these parameters according to another plane than the transversal one. Fourth rule: the reference isodose is de®ned according to the mean central dose ± and is arbitrarily equal to 85% of this mean dose. This approach was justi®ed by a large clinical experience gained in the seventies and eighties [2]. The new approach in de®ning a reference isodose is actually to take into account the target volume [3]. In brachytherapy as in external beam irradiation, the concepts of gross tumor volume (GTV) clinical tumor volume (CTV) and planned target volume (PTV) could be entirely applied. So, the de®nition of the best reference isodose is the one that covers the target volume and exclude the largest volume of normal tissue. One critical issue is to de®ne what is the PTV for each localization treated with an implant. For example, Pieters et al. [4] de®ned the PTV of a breast tumor with a focally positive margin as the tumor bed plus a margin of 2 Radiotherapy and Oncology 58 (2001) 5±6