ECC 2015
DDFMISO trial confirms significant negative prognostic value of tumour hypoxia
Prof Michael Baumann - University Hospital Carl Gustav Carus, Dresden, Germany
You’ve been looking at tumour hypoxia and specifically a new imaging modality in this study that you’ve reported here in a late breaking abstract, here in Vienna. Could you tell me about this study, what were you trying to do here?
Yes, what we did is we took some pre-clinical data which basically shows that hypoxic cells are very radio-resistant, well known, then it’s also known that hypoxia in tumours is correlated with radiosensitivity of the tumour, that was known. We did preclinical trials showing that hypoxia measured during fractionated radiotherapy is obviously a very good marker from pre-clinical experience and we put that in a prospective clinical trial.
Now, why did you look particularly at head and neck cancer?
Because the preclinical data is on head and neck cancer and so head and neck cancer is a very good model to investigate this.
So what did you do in the study?
What we did is we used a PET tracer, misonidazole, which is specific for hypoxia. We did that before treatment, we did it after one week of fractionated radiochemotherapy, two weeks and four weeks. And what we found is that the two week time point in the exploratory cohort and in the validation cohort is the best discriminator. If there is residual hypoxia at two weeks patients do unfortunately very badly. If hypoxia resolves in the first two weeks during treatment patients do very well.
Now, how would you be able to use this clinically? Can you guide and specifically tailor the conformal nature of the radiation?
What we would do now in a prospective randomised interventional trial is we would measure all patients recruited, they must be HPV negative because we know that’s another very important parameter, HPV negative patients, and after two weeks in treatment we do the hypoxia probe compare it to the hypoxia probe before treatment and if the patient is poor risk because there is residual hypoxia we will randomise the patient to dose escalation or standard dose. That has been done also in a preclinical trial and showed that, at least in the preclinical setting, that can increase the local control rate and the outcome in patients which are poor risk, which have hypoxia left.
Are the tumours homogeneous in this matter?
No, they are not but in this trial, because we have some indication also from laboratory research, that if a tumour has hypoxic areas that this is also meaningful for the areas which are outside hypoxia. So it seems to be a link between hypoxia and the tumour resistance per se. We are currently working in translational settings what does it mean. We have, for example, some preliminary indication that it might be higher stemness, a higher amount of stem cells in such tumours. So what we will do is we will give higher dose with very advanced radiotherapy to the whole tumour in the poor risk patients.
How do you envisage this being used as imaging modality in standard therapy and how could it influence therapy?
If this trial is positive then I think it can be done in many departments over the world using exactly that kind of testing. But what we do in parallel studies is to see whether we can develop other, maybe easier, ways to determine hypoxia during treatment. There are alternative ways like profiles, mRNA profiles, but they do not correlate one to one with what we see in imaging. So that needs work. But if we have this hypoxia probe, this PET imaging, that would be a fantastic way already on the way to biologically personalise radiation therapy.
And you’ve used it in head and neck cancer, could it apply to other tumours?
Yes, I think we actually found something which goes far beyond this specific trial. As I told you, the measurements before treatment of hypoxia are only a relatively week prognosticator of what’s going to happen. Measurements at two weeks during treatment are a very good discriminator. So what does it tell us about biomarkers? Maybe we have to look more on biomarkers during treatment and that’s a very important question for the whole field of personalised oncology.
Could you summarise, then, what clinicians should take away from these findings so far?
Today we know hypoxia is a very bad prognosticator for treatment outcome if measured by PET, so that can be taken as a clinical fact. Second, there is good evidence now that determination of hypoxia during treatment is more valuable as a predictive value than hypoxia before treatment. Interventional data we need to wait for the results of the trial and for the general field of oncology we need to look at biomarkers during treatment and not only before treatment.
Michael, thank you very much.
You’re welcome.
