ctDNA Could Aid in Treatment Escalation and De-escalation Decisions in Kidney Cancer

Utilizing circulating tumor DNA (ctDNA) may provide valuable insight into when to escalate vs de-escalate treatment for patients with renal cell carcinoma (RCC), but current assays are not quite ready for primetime use, according to Alan Tan, MD.

“We don’t think the sensitivity of these commercial tests are good enough, currently,” Tan explained. “I think the field is moving towards more sensitive testing, deeper sequencing, and using other technologies like methylation to uncover that.”

In an interview with OncLive®, Tan discussed the current utility and benefit of ctDNA in kidney cancer and highlighted how to integrate its use into the community setting. Tan is an assistant professor in the Department of Internal Medicine and director of GU Medical Oncology at Rush University in Chicago, Illinois.

OncLive: How is ctDNA currently utilized in the treatment of patients with cancer, specifically kidney cancer? Where does that fit into the treatment armamentarium?

Tan: ctDNA is the buzzword these days, especially in certain cancer types. It started with colorectal cancer, where we were able to look for molecular residual disease to help people understand if patients are at high risk for metastatic recurrence; it’s the premise behind getting adjuvant chemotherapy or immunotherapy in different cancers. In our field, genitourinary oncology, we have some preliminary compelling evidence in bladder cancer.

If you look at ctDNA, you could stratify patients who are the highest risk for metastatic recurrence, and maybe give them adjuvant chemotherapy. We have some ongoing clinical trials, like the IMvigor011 trial [NCT04660344], as well as another one with Alliance that’s going to be coming up next year, too. There’s a high unmet need to find these patients who need treatment and de-escalate the patients that are already cured with surgery.

In kidney cancer, however, that’s difficult because if you look at the spectrum of solid tumors, kidney cancer is one that is considered a low shedder. Therefore, we’re not so sure that ctDNA is ready for primetime to select patients that are at the highest risk for recurrence.

What preliminary evidence has been seen with the use of ctDNA detection in patients with kidney cancer?

There was an ESMO poster [presented by] urologists who compiled their real-world evidence [with a ctDNA assay]. It did select some patients who were at higher risk, but it also did miss a significant number of patients, and you don’t want to under treat. Looking at KEYNOTE-564 [NCT03142334], which is the adjuvant pembrolizumab [Keytruda] study now showing overall survival benefit, we don't want to forego treatment either in this population.

I think [we have] preliminary evidence that maybe we can start to select some patients, but I don’t think it can be applied to real world use just yet. Where I see the utility of this is in the metastatic setting because these patients usually have a higher volume tumor and shed a significant volume of ctDNA, especially when you have metastasis in organs like the bones or the liver. Sometimes you’ll see high shedding, especially in non–clear cell kidney cancer.

I have an investigator-initiated trial that I just presented at our cooperative group alliance where we can longitudinally track ctDNA. If we can clear ctDNA, maybe we can de-escalate some of these patients from some of these treatments that can be toxic, like TKIs.

There was a recent study presented at the 2023 ESMO Congress called TIDE-A [NCT04698213] that didn’t use a biomarker [to guide treatment], they instead used radiographic partial response and stable disease. What I proposed is that we use ctDNA clearance, which gives us a better understanding of if we can detect the disease at a molecular level. Maybe we can de-escalate some patients if they convert to negativity. They then can resume treatment upon early biochemical recurrence.

How have you integrated this into your own practice?

We do have a commercially available test called the Signatera™ assay by Natera. Although I can’t really speak to if it’s covered by payers currently, they’re allowing us to use this to [establish a] comfort level [in] different tumor types. I am doing it across different spectrums of GU oncology and diseases.

What I’m seeing right now is, for the most part, [ctDNA is] super informative as far as helping guide treatment responses, if patients are responding, or getting an early understanding of whether a patient is not responding to treatment. Why are we going to continue a treatment that we know is not working? We can get a preview of what’s to come on the scans. We still should decide [on treatment] based on scans first. Maybe we might get a scan earlier if we find that the trajectory [of ctDNA] is going up.

How might community oncologists use this in their own practice?

At least in CRC, it’s already being used to make treatment decisions. I wouldn’t say it’s ready for primetime to make drastic decisions yet without a prospective clinical trial validating it, but I will say it is quite promising. In fact, Medicare seems to be allowing physicians to use this when they’re tracking patients who are on immunotherapy. It’s really informative, especially when immunotherapy sometimes takes weeks to months to start to kick in [because it may indicate that] we shouldn’t give up on the immunotherapy just yet.

It’s also interesting to follow the kinetics of the ctDNA. In the beginning of using immunotherapy, you might see an uptrend [in ctDNA], and I wouldn’t worry too much about an uptrend in the beginning, because maybe 3 months later, after starting, you could have that benefit kick in; we know that’s how the immune response happens.

What main message would you like to impart to your colleagues?

The main take home message here today is not to discount kidney cancer as a ctDNA shedding cancer. It surely does shed if you have enough tumor volume, and the kinetics are highly correlative with what you see clinically. If a patient has a downtrend [in ctDNA], most likely their scans are going to be good, and the opposite is true, as well. It’s not something that’s ready for primetime for everybody right now, but I think this is the future.

One thing that we struggle with is diagnosing kidney cancer in the beginning. A lot of times these patients present with hematuria, or bleeding in the urine, and maybe large kidney masses. Sometimes these patients need a diagnosis quickly, and it’s difficult with invasive procedures to get that.

We have been using urinary track cell-free DNA to help with diagnosis. I wouldn’t say that it’s quite as sensitive for all kidney cancers, but maybe if you have a large, invasive kidney cancer that’s also presenting with blood in the urine, we can pick up mutations that are classic for kidney cancer, like in the von Hippel Lindau [VHL] gene. If you have a kidney mass in the setting or a VHL mutation in the urine, the correlation is high that it’s going to be a clear cell kidney cancer. We do have some evidence that this may be helpful, but probably not as sensitive as we would like it to be for all patients.