For the First Time, an In Vivo Laboratory Test on a Fish Has Made It Possible to Personalise Treatment for a Cancer Patient

The patient, aged 37, had a gliosarcoma – an extremely aggressive and rare form of brain cancer. He underwent two surgeries, but the cancer, revealed by symptoms such as aphasia (inability to speak) and weakness in one arm, returned. What followed was reported in Communications Medicine, a scientific journal of the Nature group.

It was at the time of the third surgery that Daniela Garcez, co-author of the article and, at the time, a neuro-oncologist at the Champalimaud Foundation (CF) Neuro-Oncology Unit in Lisbon, was faced with the task of choosing among 13 potential off-label chemotherapy treatments based on the tumour's genetic profile and literature. How could she decide which one would be the most effective?

The neuro-oncologist then decided to seek help from the CF's Cancer Development and Innate Immune System Evasion Lab, led by Rita Fior, lead author of the case study now published.

For several years, Rita Fior's team has been developing an in vivo test on zebrafish embryos called the “zAvatar-test”, in which these animals are literally used as avatars, or alter egos, of cancer patients.

To perform the test, the researchers inject the fish with a sample of the patient's tumour, generating the zAvatars. The idea is that these zAvatars then be used to test the effectiveness of the various therapies available for that patient to choose the best one – thus avoiding, in a personalised way, the administration of ineffective treatments and unnecessary toxic effects. But the test is not yet at the clinical stage.

Thirteen therapeutic options

What the researchers have been studying, for now, is the power of the zAvatar-test to predict, for a given patient, the outcome of a treatment that has already been administered to that patient. The test itself has been put to the test, in this retrospective way, in groups of patients with colorectal and breast cancer. But until now, it had never been used by an oncologist to make a therapeutic choice.

“Until now, we had only done observational studies”, explains Rita Fior. “We tested the zAvatar model by treating the zAvatars with exactly the same therapy as the patient was receiving, to see if we could predict what would actually happen to the patient.” The predictive power of the zAvatar test proved to be close to 90% in the cancers tested.

For the patient with gliosarcoma, the approach was different: the zAvatars were treated with radiotherapy (like the patient, who was being treated with focal radiotherapy) in combination with each one of the 13 potential therapies. Daniela Garcez ultimately chose the combination of radiotherapy with the chemotherapy that presented the best tolerance profile of the two most effective chemotherapies indicated by the test. She feared excessive toxicity for the patient if the more aggressive of the two were administered.

“The zAvatar-test should not override clinical judgement; it must be used to help the clinician”, says Rita Fior. “Several factors must be considered – genetic tests, clinical aspects, and the zAvatar test – to determine the therapeutic choice.” The oncologists should always have the final say, because "they are the ones who know the patient”. The final decision has to be a collaboration between the doctor and the test.

Unfortunately, the patient with glioblastoma eventually died, but the use of the zAvatar-test stabilised him and increased his survival after the third surgery. “The patient recovered from aphasia and improved his arm strength, remaining stable for two months during a very advanced stage of the disease, suggesting a benefit from this treatment during that period”, the authors write in their paper.

The case of this young patient is an isolated one, motivated by his age, the aggressiveness of the disease and a variety of possible therapies, but without a formal indication for this tumour.

Power of the test in ovarian cancer

In another paper, published almost simultaneously (on 30/12/2025) in the scientific journal Cell Reports Medicine, Rita Fior's team – with colleagues from several CF medical units and other institutions in Portugal, France and Switzerland – tested retrospectively, for the first time, the predictive power of the zAvatar-test in patients with ovarian cancer. This is also an aggressive type of cancer, which usually only becomes symptomatic at advanced stages.

The aim of this observational study was, once more, to determine the extent to which the zAvatar-test result was predictive of what actually happened to a given patient under a given treatment – and not to use the “best” choice of the test to guide treatment. In order to take this latter step, clinical trials are already open in ovarian and breast cancer.

“In our study”, explains Marta Estrada, from Rita Fior's group and first author of the paper, “we demonstrated for the first time, in a group of 32 patients”, that the zAvatar-test has a high predictive power of what will actually happen. Zebrafish embryos were injected with samples of the patients' tumours and subjected to the same treatment as was the patient who had provided the sample. The test was able to correctly predict the patients' response to therapy in 91% of cases.

In ovarian cancer, tumours often develop multi-drug resistance. “The test also made it possible to identify patients with a multi-drug resistance profile to the indicated treatments”, adds Marta Estrada.

What's more, Rita Fior points out, “in this study, in a patient with a multi-resistant tumour, we were able to show that venetoclax, a drug used in the treatment of chronic lymphocytic leukaemia, was able to make the tumour sensitive to chemotherapy.”

Venetoclax is a molecule that inhibits a protein which, in turn, inhibits apoptosis, or programmed cell death, and whose levels are often elevated in multi-resistant tumours. “We demonstrated that, in fact, the combination of venetoclax and chemotherapy was able to induce the death of cancer cells – something that had not been observed without venetoclax”, explains Marta Estrada.

Bringing zAvatars to the clinic

For the zAvatar-test to enter oncological practice, it will be necessary, as already mentioned, to conduct randomised clinical trials.

This is different from observational, retrospective studies. The aim of a randomised clinical trial is to show that if the zAvatar-test is used to guide therapy, it can actually help oncologists make significantly more adequate choices than if they were to make the decision alone.

“It's about comparing two groups of patients: one group in which the doctor chooses the therapy and another group in which the zAvatar-test chooses the therapy [with the doctor's support] – with a view to demonstrating that it is better to have a test than not to have one”, Rita Fior points out.

Her team is currently recruiting participants with ovarian and breast cancer for randomized trials. Researchers are also considering a trial for colorectal cancer, where the zAvatar-test has been shown to have a predictive power similar to that for breast and ovarian cancers.

Original Communications Medicine paper & original Cell Reports Medicine paper.
 

Text by Ana Gerschenfeld, Health&Science Writer of the Champalimaud Foundation.