Concept for individualized patient allocation: ReCompare—remote comparison of particle and photon treatment plans

Currently, the available clinical and radiobiological data for particle irradiation are for most cases insufficient to exclusively base a treatment decision on them [6, 7]. Instead, patient allocation that relies on patient-specific treatment plan comparison appears to be more robust [8] and is therefore the basis for our concept, which is explained in what follows.

The (initial) treatment decision for most patients will be made at a non-particle radiotherapy institution that lacks the competence of particle therapy planning. Nevertheless, radiation oncologists at non-particle radiotherapy institutions are clearly qualified to identify treatment plans that do not meet the criteria of the prescription (e.g., do not deliver the required dose to the tumor at acceptable levels of expected complications). For such cases, experienced personnel at the particle therapy center generate a particle therapy plan on the basis of the patient data accumulated at the non-particle radiotherapy institution, but with a different planning target volume (PTV) that considers the uncertainties at the particle therapy center. After receiving this particle treatment plan the non-particle radiotherapy institution performs a comparison with its own state-of-the-art photon plan and may finally come up with an individualized treatment decision whether to allocate the patient to a treatment with either photon or particle therapy. Note, this concept implies the important aspect of cooperation and exchange between a non-particle radiotherapy institution and a particle therapy center, which is an integral part of a well-founded decision on the different treatment options.

The workflow (depicted schematically in Figure 1) includes the preparation and transmission of patient and treatment plan data [steps 2 to 5]. The exchange is realized by the developed software tool ReCompare (REmote COMparison of PARticlE and photon treatment plans)—cf. the manuscript (Lühr et al.: Implementation of ReCompare – remote comparison of particle and photon treatment plans, in preparation) for implementation details. The preparation of the photon or particle treatment plan as well as the (final) plan comparison are performed at each institution with the same treatment planning system that is routinely used for this purpose to integrate the exchange process into clinical workflow. Patient data acquired during the application of ReCompare, i.e., pairs of pseudonymized conventional and particle therapy plans, are available from the database. They may be subject to later internal comparison for quality assurance as well as serve as a valuable data basis for clinical research regarding different irradiation modalities. The exchange, comparison, and optional scientific analysis of patient data require the informed consent of the patient. A document for this purpose—containing information for patients on the data handling—will be provided to the external non-particle radiotherapy institutions, which considers the applicable law and regulations, in particular the (German) data protection law.

The remote treatment plan exchange software ReCompare consists of three main components (as schematically depicted in Figure 2): (1) a local client at (each of) the non-particle radiotherapy institution(s), (2) one client at the particle therapy center, and (3) one central server module with a database.

The communication and data transfer between clients and the server are performed based on the Hypertext Transfer Protocol Secure (HTTPS) [9] to ensure secure data transmission. The server communicates via the internet with clients at external non-particle radiotherapy institutions and via a local network connection with the client of the particle therapy center. We preferred HTTPS to other secure protocols (e.g., File Transfer Protocol over Transport Layer Security [FTPS]) for two reasons: first, other protocols than HTTPS are not allowed by many institutions and second, HTTPS allows performing the entire communication (e.g., database requests) additionally to the data transfer through one interface. In general, individual processing steps during the plan exchange process are checked for consistency and the exchange is terminated in the case that a test fails.

A complete plan data set consists of data files in Digital Imaging and Communications in Medicine (DICOM) format [9] containing the patient computed tomography (CT), the radiation therapy (RT) structure set, the RT plan, and the RT dose distribution. Before uploading a photon plan, the files in DICOM format are (i) tested for completeness and consistency, (ii) pseudonymized, while (iii) a subset of the pseudonymized patient data (e.g., patient ID and birth date) is—after encryption using the Advanced Encryption Standard (AES) [10]—stored in a file. In contrast to fully anonymized data sets, in which all patient identifying fields are replaced, pseudonymization preserves some less identifying patient information (e.g., patient age and radiotherapy center) that might be of relevance for the treatment process. The pseudonymization has to preserve the intrinsic relations between the DICOM files, i.e., replacing the intrinsic referencing identifiers with pseudonyms in a consistent way. Otherwise, a subsequent DICOM import would fail, since the set of DICOM files could not be recognized to belong to one treatment plan. Due to the procedures in (ii) and (iii) only the non-particle radiotherapy institution can identify the patient and access patient-identifying data (e.g., name and date of birth) with their local encryption key. The checks in (i) intend to prevent faulty operation (e.g., selected patient case already exists) and to test consistency of the data (e.g., selected DICOM files do not belong together or do not form a complete plan) as well as the successful completion of processing steps (e.g., pseudonymization or writing to the database). During the initial upload process of a photon plan, the user can provide additional metadata, such as tumor location and fractionation scheme, which are then available to the particle therapy center for the purpose of treatment planning.

The final plan comparison takes place at the non-particle radiotherapy institution where the attending physician is responsible for the treatment decision. For this purpose the ReCompare software writes the dose distribution of a particle plan into a template file for a photon dose distribution in DICOM format that can be imported into photon treatment planning systems. Also during the particle plan upload, the staff at the particle therapy center may add supplementary notes regarding, e.g., the robustness of the particle plan, that are of relevance for a treatment decision. After the download of the particle plan, which is based on the same CT as the photon plan, the comparison of the dose distributions takes place in the standard treatment planning system of the non-particle radiotherapy institution and according to the procedure of that institution, which may base, e.g., on a dose volume histogram and a slice-by-slice dose level evaluation.

ReCompare fits seamlessly in the daily clinical workflow by being compatible with a number of standard treatment planning systems. These systems support planning for photon and/or proton irradiation while compatibility with planning systems capable of ions heavier than protons is under development. The configuration of the client software is kept simple and requires minimal effort by all users. The installation of the server part has to be performed in accordance to relevant requirements for protecting patient data in that country. As an example, the design of the server installation at our facility is discussed in the manuscript (Lühr et al.: Implementation of ReCompare – remote comparison of particle and photon treatment plans, in preparation).