ACTIN
For patients with metastatic cancer, multiple systemic treatment options may be available at a certain moment in time. In treatment decision making, the oncologist and patient discuss the treatment options including a prediction of expected treatment benefit for the specific patient, and weigh the benefit against potential negative side-effects.
Within ACTIN (Algorithmic Cancer Treatment Initiative), the aim is to develop a system to support oncologists and patients in treatment decision making. The system is developed as a suite of algorithms that identify all possible treatment options (standard-of-care and experimental) for the specific patient and predict the personal expected benefit of treatments (in terms of survival) based on published literature and real world treatment efficacy evidence.
ACTIN is developed in close cooperation with hospitals, to tailor the output of the algorithms to meet clinical needs and ensure tight integration into the hospital infrastructure to promote seamless usage by physicians.
What was the reason this project was started?
Results displayed within Hartwig’s whole genome sequencing diagnostic product (OncoAct) can be indicative of experimental or standard-of-care treatments. However, other clinical factors are relevant for treatment decision making as well, such as radiology results, pathology results and lab results. In addition, interpretation of molecular results can be challenging, both from a biological or analytical point of view and considering new evidence and indications are evolving rapidly (for example as described in this article).
The ACTIN initiative started in summer 2021 as a collaboration between Erasmus MC and Hartwig Medical Foundation, with the aim to build a suite of algorithms to address these challenges.


What are the key principles of the ACTIN system?
For the development of the ACTIN system, the following principles are adhered to:
- The system does not intend to replace treatment decision making. ACTIN only intends to provide physicians and patients with available treatment options and personalized efficacy predictions and thereby providing more information for making the treatment decision.
- ACTIN aims to integrate seamlessly with hospital infrastructure, such that it will be available for physicians with minimal effort required.
All of ACTIN’s algorithms are open-source and its infrastructure is considered to be part of a wider ecosystem of infrastructure in the healthcare space. Cooperation with researchers generating new evidence and with technical infrastructure initiatives to leverage efforts in this space is considered part of ACTIN’s scope.
What has been done so far?
At Erasmus MC, the ACTIN development started within the setting of the early phase clinical trial team. This team identifies early clinical trial treatment options for patients who have no standard-of-care options available. Here, a system was built to ingest all relevant patient and trial data from the hospital systems and automatically determine eligibility for applicable clinical trials running at Erasmus MC. This pilot has been completed, and the system is in process of being deployed for all phase1 patients at Erasmus MC.
In 2024, NKI-AvL implemented the ACTIN trial matching algorithm within the context of lung cancer. This algorithm is now also in process of being deployed for all lung cancer patients at NKI-AvL.
In 2023, together with Erasmus MC the development of an algorithm started to predict the personal expected treatment benefit for standard-of-care treatment options for a specific patient. The system is initially being developed for metastatic colorectal cancer patients who are eligible for first-line treatment in a palliative setting. The development of this algorithm is currently work-in-progress, with UMC Utrecht joining this initiative as of the start of 2025. In this setting, the data for every patient that is evaluated by this algorithm will be used to improve treatment benefit prediction for the next patient.


What are the plans for ACTIN going forward?
Presently, ACTIN is still in development. Together with participating hospitals and various other partners Hartwig is working on the following themes:
- Making trial eligibility identification available to more patients and covering more trials.
- Developing an accurate and clinically useful personalized treatment benefit prediction algorithm for treatment options within standard-of-care.
- Contributing to a larger ecosystem of technology which enables easier reuse of data and enables applications such as ACTIN to be integrated into existing health care platforms used by physicians in hospitals.
If the practitioner doesn’t know where the primary tumor is located, they cannot offer a targeted treatment plan, and nor can they give the patient an estimate of their life expectancy.