Real-time analysis of anti-tumor activity and toxicity outcomes during phase I dose escalation trials could improve dose selection: a PARP inhibitors case study

Finding the right amount of medicine is one of the most delicate tasks in treating cancer. For decades, the standard approach for testing new drugs was to give patients increasingly higher doses until the side effects became too severe to ignore. This method, designed for older, harsher chemotherapy drugs, assumed that a stronger dose always meant a stronger attack on the tumor, and that the highest dose a patient could survive was the best one to use. However, modern cancer treatments work differently. These newer agents are like precision tools designed to target specific weaknesses in cancer cells while leaving healthy tissue alone. Because they work through a different mechanism, the point where they stop getting more effective is often much lower than the point where they become dangerously toxic. If researchers stick to the old rule of "more is better," they risk giving patients doses that are far higher than necessary, causing unnecessary suffering without adding any extra benefit.

A team of researchers from the Medicines Evaluation Board and Radboud University Medical Center decided to test whether looking at both the benefits and the risks together could change how these doses are chosen. They focused on a specific group of drugs called PARP inhibitors, which are used to treat certain types of cancer and have been approved for use in Europe. These drugs are known to cause significant side effects, often forcing patients to lower their dose or stop treatment entirely. The researchers wanted to see if they could use data from the very first stages of clinical trials to map out exactly how the drugs worked. Instead of just watching for toxicity, they also tracked how well the tumors shrank at each dose level. They used a straightforward statistical method that simply arranges the data to show a clear, rising trend, allowing them to draw a picture of the relationship between the amount of medicine given and the results it produced.

The researchers gathered information from the early clinical trials of four different PARP inhibitors: olaparib, rucaparib, niraparib, and talazoparib. In these initial studies, small groups of patients received different doses of the drug, starting low and going higher. The researchers collected the specific numbers for how many patients experienced severe side effects and how many saw their tumors shrink at each of these dose levels. They then plotted these results to create curves that showed the probability of a drug working and the probability of it causing harm as the dose increased. This approach allowed them to see the full shape of the drug's performance, rather than just looking at a single point where the side effects became too much.

The results revealed a pattern that challenges the traditional way of selecting doses. For every drug they analyzed, the ability to shrink tumors reached a peak and then flattened out, forming a plateau. Crucially, this plateau began at a dose level that was lower than the maximum dose the patients could tolerate. In other words, increasing the dose beyond a certain point did not make the drug work any better, but it did increase the risk of severe side effects. For three of the four drugs, the dose officially recommended for later stages of testing was the maximum tolerated dose, which sat right in the zone where the drug was no longer getting more effective but was becoming more dangerous. Only for one drug, olaparib, was a lower dose chosen as the recommendation, and this happened to align perfectly with the point where the drug's effectiveness stopped increasing.

One specific case highlighted the importance of this finding. For the drug niraparib, the recommended dose was set at 300 mg once daily. However, after the drug was approved, a follow-up study showed that patients who had their dose reduced to 200 mg or even 100 mg had the same survival rates as those on the higher dose. This led to an official change in the prescribing information to lower the recommended dose for certain patients. The researchers noted that this real-world adjustment confirmed what their analysis of the early data had already suggested: the optimal dose was lower than the maximum tolerated dose. For another drug, rucaparib, no maximum tolerated dose was even found during the initial trials because the side effects did not follow the expected pattern, yet the analysis still showed that the drug's effectiveness had already peaked at lower levels.

The study suggests that relying solely on toxicity to decide the dose for new targeted cancer therapies may be leading to the selection of doses that are too high. By looking at the entire picture of how the drug behaves—seeing where the benefits stop growing and where the risks start rising—researchers can identify a "sweet spot" that offers the best balance. This approach does not require complex new technologies or massive amounts of data; it simply requires paying attention to the effectiveness numbers that are often collected but overlooked during the early phases of testing. The authors argue that this method provides a clearer view of the therapeutic window, the range of doses where a drug is both safe and effective.

This work serves as a reminder that the rules developed for older chemotherapy drugs do not necessarily apply to modern precision medicines. The study indicates that even with the limited data available from early trials, it is possible to map out the dose-response relationship and find a more suitable dose for patients. By shifting the focus from simply finding the highest dose a patient can survive to finding the dose that offers the best biological response, the medical community can potentially reduce the burden of side effects for patients while maintaining the power to fight the cancer. The researchers conclude that this perspective can support better decision-making for regulators and drug developers, ensuring that the next generation of cancer treatments is optimized for the patient's well-being from the very beginning.