Arlene O. Siefker-Radtke, MD: We've all seen and heard about the concepts of pruning, where prior treatments may alter the environment, kill off certain populations of tumor cells, and leave additional tumor cells or mutations behind, as we've seen with increases in tumor mutation burden following chemotherapy.… [I] like getting that biopsy before a patient goes on treatment whenever it's feasible, but to do a good mutation test on it, you need a good tissue sample. And that is something that can be a challenge even at a place like MD Anderson, where our interventional radiologists will biopsy early and often, and even those lesions that are hard to get. We do sometimes run into that challenge where there's not enough tissue on a biopsy because you need a good core needle. You're not going to get enough of the tumor on a fine-needle aspirate or smaller, like if you see a 23[-gauge] French core-needle biopsy that's heading a little more toward an FNA [fine needle aspiration].

These other panels can be helpful and very supplemental to what we need to know. You mentioned looking at the primary, and I'll take a muscle-invasive primary if that's what I have available for testing. And then there's the liquid biopsies that we're doing in the circulation, where we're able to find the mutations present through these circulating DNA-type assays. But the challenges could be germline alterations that are present, and we don't always pick up the mutations. And in fact, there was an abstract that we presented on the Ortho Fitbit data GEO ASCO [Genitourinary Cancers Symposium American Society of Clinical Oncology] comparing liquid biopsies with tissue biopsies. And when we looked at patients who had the liquid biopsy and had FGF [fibroblast growth factor] receptor alterations present, we saw the response rate in over 50% of patients. It did tend to predict for those where FGF may be driving the tumor. But even in the group of patients where we could not detect the tumor or detect the mutation in the circulation, we still saw a response rate in about 30% of patients who had FGF alterations in the tissue. It may require that multitargeted approach where we look at tissue, [and] we look at circulating DNA to try to predict the best patient for selection for the appropriate therapies. I'm just curious, Scott, do you check patients when they have metastatic disease? Do you check any patients in the neoadjuvant or preoperative or perioperative space?

Scott T. Tagawa, MD, MS, FACP: We have dual roles in terms of clinical care as well as research, so in terms of pure clinical care, I would say what is prime time is testing in the setting of metastatic disease. We often will get as much tissue as we can at every single time point that's available. For instance, we will assay their diagnostic TRB [transurethral resection of a bladder]. If they have residual disease, we may be looking mostly for research purposes at what is left over if they have anything at the time of cystectomy. And if they have a recurrence, we will do additional testing there, which especially at that third time point, we’ll sort of dual-role. That is the clinically relevant one, at least the most relevant one, but we do tissue serial testing in our patients that are getting tissue-sampled as part of standard of care. And then obviously, as part of some clinical trials, we'll occasionally do some research biology. We tend to assay frequently, but that is in part because we are both looking to help the patient sitting in front of us, but also future patients through research.

Arlene O. Siefker-Radtke, MD: And that's a great point for our colleagues in the community to hear: There are a lot of trials of targeted agents across the spectrum of bladder cancer even in earlier stage tumors. And if you look at FGFR3 specifically, it's present in over 60% of superficial bladder tumors and still present in about 15 to 20% of metastatic bladder tumors and about 35% of metastatic upper tract tumors.

Transcript edited for clarity.