Title
Benedikt
STATE OF THE ART
FUNCTIONAL AND MOLECULAR IMAGING BASED DOSE PAINTING
RESPONSE-ADAPTIVE RADIOTHERAPY
Dose painting
Dose painting is the prescription of a non-uniform, heterogeneous radiation dose distribution to a target volume based on functional or molecular imaging (missing citation). The procedure is mainly motivated by three hypotheses.
Highly radioresistant fractions of the tumour volume are not eradicated by a safely deliverable uniform dose distribution and can be the cause for local recurrences.
Molecular and functional imaging like positron emission tomography (PET) allows spatio-temporal mapping of the tumour volume and provides information about the local relative radioresistance.
Modern radiotherapy planning and delivery systems are able to deliver a graded boost to the high-risk sub-volumes, potentially increasing the tumour control probability (TCP).
Uniform, homogeneous dose distributions within the tumour volume include the risk of overdosing parts of the target volume while still failing to eradicate the high-risk sub-volumes (missing citation). Therefore, a heterogeneous dose distribution taking into account variations within the tumor volume appears desirable. There are two strategies to realise dose painting: sub-volume boosting (missing citation) and dose painting by numbers (missing citation).
Sub-volume boosting applies an additional dose to one or more smaller targets within the primary target. The sub-volumes are defined by image segmentation on the basis of quantitative information. In theory, this is a binary technique using two discrete dose levels, one for the primary target and a higher one for the sub-volumes.
Dose painting by numbers is a voxel-based technique and aims to apply adapted dose levels based on a mathematical transformation of the image intensity of individual pixels. Ideally, an individual dose is applied to each voxel. This method provides a higher degree of biological conformity, but also places higher demands on the imaging and radiotherapy systems.
Combinations of the two strategies define several sub-volumes, each assigned to a different, individual dose. The functional images required to obtain information about potential high-risk structures can be acquired using combinations of CT and PET scans. Various biomarkers, each connected to particular PET tracers, have been identified to be related directly to the therapeutic response and an increased risk of recurrence. The three most commonly considered biological characteristics are glycolytic metabolism, proliferation and hypoxia . Although there are favorable tracers for each of these tissue properties, it has to be noted that any tracer usually depends not solely on one but on several factors.
Imaging biomarkers for the risk of recurrence after radiotherapy
Glycolytic metabolism, clonogen density
Clonogen density, the overall amount of clonogens within a tumour and the rate of their glycolytic metabolism are the most intuitive criteria for the prescribed dose in radiotherapy. The more clonogens there are, the more dose is required. Histological tumour specimens show a variation of clonogen density throughout the tumour volume. It can be visualised with 〖^18〗F-Fluordesoxyglucose (FDG) PET. In addition to its high signal-to-noise ratio, further advantages of FDG are its easy availability and its common use in clinical routine and research. Many departments already gathered the extensive experience required to make FDG uptake a feasible and reliable biomarker (missing citation); (missing citation); (missing citation); (missing citation). However, the standardized uptake value (SUV) is not solely dependent on clonogen density but also on other factors such as tumour perfusion, rate of glycolysis, proliferation, inflammation and hypoxia (missing citation); (missing citation). So far, there is only limited data available on alternative tracers such as amino-acids (e.g. MET) (missing citation), fluoroethyl-L-tyrosine (FET) or 〖^11〗C-choline (missing citation), which makes FDG the most profound tracer regarding clonogen density.
Proliferation
Another important criterion for the TCP is the tumour cell proliferation. In today’s radiotherapy planning it is already taken into account in terms of biological effectiveness of a prescribed dose. An overall increased treatment time demands a higher overall dose to compensate for the new tumour cells emerging during treatment (missing citation). Experience in radiotherapy associates higher proliferation rates with a worse therapy outcome. Higher radiation doses are required to control rapidly proliferating tumours. Therefore, varying proliferation rates within a tumour volume are a motivating factor for dose painting on the basis of proliferation biomarkers.
Various PET tracers have been associated with tumour proliferation. Research had initially been centered on 〖^11〗C-labeled thymidine and several radiolabeled halogenated pyrimidine deoxynucleosides, until studies revealed complications in image interpretation caused by rapid biodegradation of the tracers and resulting radiolabelled metabolites within the blood (missing citation); (missing citation). Current interest focuses on 3’-deoxy-3’-[〖^18〗F]fluorothymidine (FLT), another radiofluorinated thymidine analog. FLT is a marker taken up by the cells and phosphorylated by thymidine kinase 1, which is an enzyme closely tied to cellular proliferation. FLT PET imaging provides a profound map of the local growth fraction of tumour cells and its SUV appears to be a promising biomarker for proliferation and cell kinetics in general (missing citation).
Some studies found that both pre-treatment cell kinetics and pre-treatment cell growth are not predictive of fractionated radiotherapy outcome, putting treatment planning based on proliferation biomarkers acquired at baseline into question (missing citation); (missing citation). However, the change in FLT status between measurements at baseline and after the start of radiotherapy does indeed seem to be a biological predictor of therapy outcome, making the proliferation rate during therapy a more important criterion than the proliferation at baseline (missing citation). Therefore, additional image acquisition during treatment is crucial for an adaptive radiotherapy based on tumour proliferation. Studies also showed a statistically significant relation between the epidermal growth factor receptor (EGFR) and a favorable therapy outcome for patients with head and neck squamous-cell carcinoma (HNSCC) . The expression of the EGFR can be PET imaged using radiolabeled tyrosine kinase inhibitors or antibodies against the extra-cellular domain of the receptor (missing citation).
Hypoxia
Hypoxia is both a consequence and a driver of malignant progression and leads to a poor therapeutic response to radiotherapy . The biological effectiveness of radiotherapy highly depends on free radicals from oxygen to effectively destroy the target cells. This is why hypoxic areas show a signficantly higher radioresistance and require an up to three times higher dose for the treatment to achieve the same level of biological effectiveness as in normally oxygenated tissue (missing citation). Adaptive dose optimisation based on a hypoxic biomarker is most likely beneficial for both boosting high-risk sub-volumes (missing citation) and maybe even de-escalating dose in tumour areas with a higher oxygenation (missing citation).
Current studies typically use the PET tracers Fluormisonidazole (FMISO) or Cu(II)-diacetyl-bis(N〖^4〗-methylthiosemicarbazone) (Cu-ATSM) to image hypoxia. Another direct approach to prove and measure hypoxia in tumours was first taken with oxygen electrodes (missing citation). However, the procedure is not feasible for radiotherapy treatment planning. Although it is difficult to prove the quantitative validity of the FMISO and Cu-ATSM SUVs, there is an abundance of research showing promising results for adaptive radiotherapy based on these uptakes as biomarkers for hypoxia. In addition to hypoxia imaging at baseline, temporal changes in oxygenation might have substantial influence on the TCP. A second or even third PET image after certain periods of treatment time might be reasonable (missing citation). Aside from FMISO and Cu-ATSM, there are other possible tracers such as FAZA, FETNIM, EF5 or EF3. Their differences in uptake characteristics and in hypoxia localisation are of special interest and need to be further investigated, in spite of (or actually even due to) the possibility of a lack in co-localisation and boost-volume definition.
Response-adaptive dose painting based on PET imaging
With available information about the local radioresistance within the tumour volume, the most general method for modelling the tumour response is to calculate the TCP at voxel level and then to integrate this function over the whole target volume (missing citation). For practical reasons, studies so far proposed various hybrid models of dose painting by numbers and sub-volume boosting, defining one or more sub-volumes with different risk- or dose-levels. A general problem is the fact that the relevant tumour phenotypes discussed earlier can not be imaged at the same time but only extended over several days in order to allow for a clearance of the different PET tracers between the scans. This complicates both the acquisition of one final image to base the treatment planning on and the calculation model itself. Apart from the usual setup uncertainties occurring in multimodality imaging, especially with individual images taken at different times, the spatial and temporal heterogeneities and changes of the biological system make the definition of a time-independent biological target volume (BTV (missing citation)) even more difficult. Nevertheless, it is possible to define a biological target volume based on pre-treatment PET imaging.
To take into account the uncertainties of the pre-treatment imaging, the dynamics of the biological system during the treatment period and the radiation-induced biological changes of the tumour, some groups proposed the concept of adaptive radiotherapy based on additional functional imaging during the treatment period . While direct monitoring of the delivered dose during treatment is not yet possible, monitoring the tumour response does indeed seem feasible via repeated PET imaging . This kind of response assessment could eventually allow for an individually adapted radiotherapy based on functional imaging during the therapeutic process.
In both pre-treatment and response-adaptive BTV definition a threshold value can help to quantify the relative changes in radiosensitivity within the tumour volume as well as the effective radiosensitivity of the patient as such. Another crucial parameter is the time window used for evaluation. Recent publications agree in a time window of one to two weeks between the start of treatment and the response assessment, although further research has yet to be conducted .
Conclusion and future prospects
Functional imaging has the potential to provide a paradigm shift in radiation treatment planning and to substantially impact clinical outcomes. With its ability to not only consider the pre-treatment conditions of patients but also their intrinsic responsiveness to the treatment, adaptive dose painting based on PET imaged biomarkers provides an important step towards biologically guided radiation therapy.
Due to the variety of possible biomarkers and the differences in their respective PET tracers, finding the most suitable combination of tracers and proposing a quantitative radiobiological model to simultaneously account for them is and will be one of the main objectives of current and future studies. An at least equally challenging task is the further development of profound and clinically feasible dose painting calculation models to adequately process the data acquired by functional imaging. To do so, further research and clinical studies are indispensable.
References