ESTRO 2023 Biology Track Report – Best Paper
A short interview with author Apostolos Menegakis
“Hypoxic tumour cells drive tumour relapse after radiotherapy as revealed by a novel tracing tool”
What was your motivation for initiating the study?
During my PhD and the early period of my post-doc, I worked in laboratories that had very close links with radiation oncology clinics. There, for the first time, I came to understand from the clinical perspective that tumour hypoxia constitutes an important issue that leads to therapy resistance. I was intrigued by two facts. On the one hand, hypoxia is a common feature of many solid tumours, and in the literature, patients with more hypoxic tumours (independent of the way in which that is assessed) have poorer prognoses. The latter is particularly evident in tumour types for which radiotherapy plays a critical role in their therapeutic management. On the other hand, the hypoxia problem has been known for a long time in the radiotherapy field and despite numerous efforts to target it in patients, thus far, it has not been possible to translate an efficient strategy to “tackle” hypoxia to most of the cases. Furthermore, data from radiobiological studies indicate that hypoxic cells are more radioresistant than non-hypoxic cells, which has been attributed mainly to the oxygen effect and reduced induction of DNA damage by irradiation. However, it has been extensively shown that irradiation also leads to tumour mass reoxygenation during the course of fractionated radiotherapy modules, which therefore should render previously hypoxic cells more sensitive to subsequent fractions. In a simplified approach, if the problem is the lack of oxygen at the time of irradiation, then during treatment that reoxygenates tumours, the problem should be counteracted. In addition, endeavours to dose-escalate the hypoxic volume, which directly targets hypoxic cells with hypoxia-activated pro-drugs and should reoxygenate the tumour, should have yielded excellent results.
The obvious and puzzling question then was: “Why does hypoxia remain such an important issue?” As a researcher, my principle is that the key aspect to solving a clinical problem is to understand the mechanistic background that leads to the observed phenomena Therefore, I struggled with the idea that although presumably the mechanism was known, a solution could not be found. I kept asking myself: “What are we missing here?” I am very grateful to Professor René Medema, because when I came to his lab and shared my thoughts, he offered his unconditional support to try and re-address the hypoxia problem in radiotherapy from scratch.
One thing we realised early on was that although hypoxic cell radioresistance and the correlation of hypoxia with poor prognosis were well documented, the direct evidence that hypoxic tumour cells gave rise to tumour relapse after radiotherapy was actually missing! Approaches to tag and follow hypoxic cells have proved inefficient and this has made it impossible to visualise these cells at the moment that they are hypoxic and thus to study their biological behaviour during and post-irradiation.
What are the main findings of your project?
We have teamed up with the group of Marc Vooijs in Maastricht, where they have developed a novel approach to tag and trace hypoxic cells. They have built up a hypoxia reporter based on the conditional expression of a fluorescent protein, which not only enables us to visualise hypoxic cells at any given time point, but also to trace their progeny over time. We validated the hypoxia reporter in 2D cell cultures, a 3D spheroid model, and in human tumour xenografts in a human lung adenocarcinoma cell line. We then asked the question: “Are the hypoxic cells that are present at the beginning of a radiation treatment responsible for tumour relapse after radiotherapy?” The results were impressive! Both in spheroids and in tumours, the cells that were initially hypoxic populated the regrowing spheroid and the relapsed tumour. These data provide the first direct evidence that these are the cells that drive tumour relapse after radiotherapy.
The next question was: “How is this actually happening?” We thought of the amount of DNA damage and that we would have the ability to prove their survival advantage. However, to our surprise, an interesting biological behaviour was found that seemed more complicated. These cells were merely quiescent and, importantly, they could remain quiescent for very long periods after irradiation, before they “woke up” to give rise to relapse.
What are the implications of this research?
Our findings have important implications for the translation of approaches to target hypoxic cells in the clinic. Our data suggest that the radioprotection of hypoxic cells is not only attributed to the oxygen effect but additionally to the fact that they are quiescent and not actively cycling. Therefore, the timing of any treatment intervention strategy to target these cells is critical, as they can remain arrested (and thus protected from the treatment) for prolonged periods and can later give rise to relapse. As proof of principle, we have shown that increasing the duration between fractions and aligning the subsequent fractions to the time at which these cells re-enter the cell cycle, is a good strategy to target them at least in 3D models. We aim to exploit that opportunity further in xenografted tumours. We hope that the utilisation of novel techniques will provide better insights into the biological behaviour of hypoxic cells and how they contribute to tumour relapse after radiotherapy. The latter can be the basis for the designs of innovative targeted strategies that will lead to better therapeutic outcomes for patients.
What are the future perspectives?
We have now acquired strong evidence that hypoxic cells contribute to tumour relapse after radiotherapy. However, our mechanistic understanding remains elusive. Our data suggest that radiation-induced reoxygenation does not coincide with the re-entry of hypoxic cells into the cell cycle (at least not in all cases). Instead, it seems more likely that at least some of the hypoxic cells must acquire specific molecular traits before they are enabled to drive the relapse. Understanding what drives this transition from a quiescent hypoxic tumour cell to a tumour cell able to lead the tumour relapse will be the focus of our research in the coming years. In addition, we aim to expand our observations from a lung adenocarcinoma model to more tumour types that involve radiotherapy in their therapeutic management (e.g. head and neck, glioblastoma).
Apostolos Menegakis
Post-doc in radiation biology, Department of Radiation Oncology
The Netherlands Cancer Institute (NKI)
Amsterdam, The Netherlands