We recently did a collaboration at MSK looking at the CURB trial, more specifically the non-small cell lung cancer cohort who received SBRT for oligoprogression. The key finding from that study was that the sum of longest diameters of growing lesions, which is the amount of lesions, the burden of lesions, that were growing at the time of oligoprogression was the tumour burden metric that was most strongly associated with outcomes. We used a median cut-off, the sum of longest diameters of growing lesions less than or equal to 4cm identified patients who had markedly longer progression free survival and overall survival than sum of longest diameters of growing lesions greater than 4cm.
What were the results and methodology?
In terms of actually quantifying the differences, this is a difference in progression free survival that is over 20 months difference. So why did we focus on growing lesions? Actually at the time when we thought of this secondary analysis of the CURB trial we wanted to see which tumour burdens were most prognostic of PFS and OS because typically a lot of people assume that it’s just the number of progressing lesions. So we wanted to do a systematic approach looking at not only the number of progressing lesions, the number of all lesions in patients, but look at the volume, what the sum of longest diameter of all lesions, sum of longest diameter of progressive lesions and apply univariable and multivariable analyses to objectively ascertain which is the most meaningful metric and most impactful from a prognostic standpoint.
The purpose of doing such a thing would be to in the future, if validated in future studies, stratify decision making by the sum of longest diameter of growing lesions, or the volumetric tumour burden at oligoprogression. Because if you look at our data, even just glancing at the swimmer plot, there’s a marked difference in terms of if you had a sum of longest diameter of growing lesions less than 4cm most people were certainly living beyond a year but not even progressing until more than a year after treatment, while people with greater volumetric tumour burden were progressing and dying within a year, most patients. So if a patient has a lower volumetric tumour burden of progressive disease perhaps SBRT to progressing sites – SBRT is stereotactic body radiotherapy, high dose radiation, delivered over less than or equal to five treatments – delivering that to progressive sites and continuing on current systemic therapy could be a viable option. However, patients with a greater volumetric tumour burden of progressive disease, in addition to SBRT to the progressive lesions, considering intensifying or switching systemic therapy or a trial, based on our data, would make sense and be justified.
One thing I would like to clarify, because this has come up: although we wanted to have relatively evenly balanced groups, and that’s how we came up with this cut-off of 4cm, I believe that this really occurs along a spectrum. So the lower your volumetric tumour burden and progressive lesions, I suspect that the probability of SBRT plus current systemic therapy being effective is just higher. But as you get to have a larger sum of longest diameters of growing lesions or volumetric tumour burden of growing lesions, the probability of SBRT to growing lesions alone being effective, it’s just going to progressively go down. So although to really demonstrate the relationship graphically is categorical variable, I think that this is something that will exist along a spectrum and patients will have to go over with their physicians in terms of identifying which approach is optimal for treating their oligoprogressive non-small cell lung cancer.
What are the clinical implications and what is next for this study?
For this study, the next thing we’re doing is looking at circulating tumour DNA and seeing how that correlates with both a sum of longest diameter of growing lesions as well as the volumetric tumour burden. So rather than summing up the longest diameter of each lesion in a patient and each growing lesion in a patient, actually doing a volumetric measurement. So we want to have a more comprehensive characterisation of tumour burden, particularly in these progressive lesions because technology like auto-segmentation will allow us to leverage AI to accurately quantify in the most objective means possible the tumour burden both overall but also progressive lesions. Because it seems the tumour burden of progressive lesions is the most meaningful from a prognostic standpoint for progression free survival and overall survival. It would always be great if we could have a peripheral blood assay like circulating tumour DNA measurements to be able to give patients results that could be meaningful for helping decide which therapy is optimal for them.
We’d like to validate this in a larger dataset too; we’re working on that. I should also say the CURB II trial is enrolling. Dr Tsai who was the PI of the CURB trial and she’s been working closely with us on this secondary analysis, she’s leading a randomised phase III trial that’s studying this question of specifically adding SBRT to first-line treatment or starting new second-line therapy in the setting of non-small cell oligoprogression, which of those arms is going to have the best outcomes. There are co-primary endpoints of progression free survival and overall survival
