This study was a post-hoc biomarker analysis of the CheckMate 214 randomised phase III trial. CheckMate 214 was a phase III trial that for first line advanced renal cell cancer patients compared the old standard of care, sunitinib, versus combination immunotherapy with ipilimumab and nivolumab. The previous published results from CheckMate 214 showed that combination immunotherapy with ipi/nivo had a durable overall survival benefit compared to sunitinib and previously established combination immunotherapy as a standard first-line kidney cancer regimen.

However, we have not had an available blood test that could tell us which patients are those who had the incredible long-term many-year responses to immunotherapy and which are the patients that don’t respond well to immunotherapy. In this study we explored whether a blood test for KIM-1 might fulfil that role.

What was the study design?

This study was a post-hoc biomarker analysis. We obtained banked serum from 821 patients who participated in CheckMate 214 and this represented about 75% of the overall study population. In this study we checked KIM-1 levels at baseline before treatment and three weeks into the study after a single dose of immunotherapy or three weeks of sunitinib. We compared whether patients’ outcomes were different depending on high versus low KIM-1 at the pre-treatment baseline and depending on the amount of KIM-1 change after three weeks.

What were the results of this study?

In this study, first, we found that baseline KIM-1 was elevated among this overall population of patients with renal cell carcinoma which is not surprising given that there have been several prior studies showing that serum KIM-1 is a biomarker for renal cell carcinoma disease status. This is a protein that’s shed from the cellular membrane and patients who had higher disease burden had higher levels of KIM-1 as expected. We also found that the baseline pre-treatment KIM-1 levels were a good predictor of the patient’s overall risk status. Patients with high KIM-1 did worse for progression free survival and overall survival while patients with low KIM-1 did better. This biomarker, KIM-1, was independent of either IMDC status and remained significant after adjustment for tumour burden. So it was a good baseline prognostic marker.

But what was more exciting to us was that when we looked at the change in KIM-1 after three weeks of immunotherapy, patients who had an increase versus a decrease in KIM-1 after one cycle of ipi/nivo really had quite a divergence in their clinical outcomes. For patients who had a 30% decrease in KIM-1 versus those who had a 30% increase in KIM-1, those who had a 30% decrease at three weeks, which was 31-32% of the overall population had a median progression free survival of 71 months versus only 4 months who had a KIM-1 30% increase at three weeks. So what we’re finding is that this also correlated with overall survival and it’s an early way to know how well someone is going to do on immunotherapy long-term.

Additionally we looked at whether the change in KIM-1 affected patients on the sunitinib arm and we found that the 3-week KIM-1 was not a predictor for 3-week response to sunitinib. So this effect was pretty specific at the three week mark to the immunotherapy regimen.

What is the clinical significance of these results?

Right now many, many patients with kidney cancer, close to half, are treated with combination immunotherapy in the first line. Some of these patients do great. About a third of patients have long-term remissions and even treatment-free survival. But about one in five patients have no response at all to combination immunotherapy. For those patients who don’t have a response oftentimes by the time they have their first scan these patients are having toxicity from treatment and disease growth and arguably these patients should not be continued on immunotherapy if we could have some way of knowing that they’re not going to do well on this treatment.

This blood test provides some hope that we can identify early on at three weeks which patients are going to do great and which patients would be better served by switching to an alternative strategy, one that will work better and not  have this unnecessary toxicity for a treatment that’s not going to work.

Is there anything else you would like to add?

This is a post-hoc biomarker analysis so the findings are necessarily preliminary and we are working on validating these results in several other large trials as well as starting prospective trials where we can test using KIM-1 to make management decisions for kidney cancer.