CLL-CLUE: Personalising Treatment for Europe’s Most Common Leukaemia

Chronic lymphocytic leukaemia (CLL) is the most common form of leukaemia in Europe, and despite significant medical advances, it remains incurable. While modern medicine has moved toward targeted drugs, designed to attack the molecules that help cancer grow rather than killing all fast-growing cells like chemotherapy, these therapies do not work the same way for everyone. Two people with the same disease can respond very differently to the same treatment, and hence personalised medicine is needed for more successful and efficient outcomes.

Physicians once relied on DNA markers to attempt to predict outcomes, but these traditional genetic tests fail to predict how patients respond to new targeted drugs. This gap led to the creation of Functional Precision Medicine (FPM). While traditional precision medicine looks for mistakes in a patient’s DNA, FPM looks at the behaviour of the cell – specifically the proteins and signaling pathways that act as the cell’s internal communication machinery. By observing which of these signaling pathways are turned on or off, researchers can gain a much clearer picture of how a patient’s cancer will most likely behave when it meets a specific drug.

A Breakthrough in Predicting Success

The CLL-CLUE project, an ambitious interdisciplinary consortium funded in the framework of ERA PerMed Joint Transnational Call 2020 (JTC2020), utilised FPM to improve the treatment outcome of CLL. The consortium recently reached a major milestone with a landmark study published in Clinical Cancer Research. A joint collaborative effort led by CLL-CLUE consortium members Sigrid Skånland and Tero Aittokallio used a specific targeted drug (a PI3K inhibitor) as a model to understand how these therapies change the internal communication lines, or cell signaling, within leukaemia cells.

By combining deep biological analysis with artificial intelligence, the researchers identified a ‘signature’ of 5 signaling pathway proteins phosphorylated at specific sites that could act as a predictor for drug resistance. They discovered that the model could predict whether a patient would be a “responder” (someone whose cancer shrinks or stabilises) or a “non-responder” (someone whose cancer continues to grow) with 80% accuracy before the treatment even began. Furthermore, the model reached 89% accuracy when distinguishing between those who would achieve a partial versus a complete response. This level of precision is possible because the team moved beyond static DNA markers to look at the active protein machinery inside the cell.

In patients who achieved a positive response, the team observed that the drug treatment successfully shut down specific internal signals, such as the AKT signaling pathway, which helps the cancer to survive. Simultaneously, the treatment triggered “pro-death” signals like the Bim protein, causing the leukaemia cells to self-destruct. The study further revealed that the responding patients experienced a significant increase in mature, “cytotoxic” Natural Killer (NK) cells, the body’s native defence, suggesting the treatment helps the patient’s own immune system join the fight against the cancer.

For patients categorised as non-responders to the PI3K inhibitor treatment, the researchers continued to explore alternative treatments that could work for them. A sample of each of the non-responder’s cancer cells were screened with 94 different drugs and across 87 drug combinations to find a personalised strategy that would break through the resistance and halt cancer growth. They found a particular sub-set of the non-responders that were pre-clinically sensitive to a combination of PI3K inhibitor with a BCL2 inhibitor. This result brings the CLL-CLUE consortium one step closer to their goal- that all patients will be prescribed drugs that most effectively work for them, sparing them the time, cost, and pain of ineffective treatments.