Case 2: Combination Therapy for Cis-Ineligible Bladder Cancer

EXPERT PERSPECTIVE VIRTUAL TUMOR BOARD

Arlene O. Siefker-Radtke, MD:We’ll have to see. If we did have to give chemotherapy, what if we didn’t know PD-L1 [programmed death-ligand 1] status, and they’re rapidly progressing, what chemotherapy would you choose?

Tian Zhang, MD:I would still probably go with gemcitabine and carboplatin. The carboplatin is renally dosed, so whatever creatinine they come in with that day, you calculate their creatinine clearance and you find your carboplatin dose. So I still think that a gemcitabine-based, platinum-based and combination in that setting would be my first-line go-to.

Elizabeth R. Plimack, MD:Yes, definitely if truly CIS [cisplatin]-ineligible. I’ll throw in 1 other caveat with this case is that the sort of labeling of this patient of cisplatin-ineligible is based on a calculated and estimated creatinine clearance. If this patient had ureteral obstruction, I would send them to Gordon for perhaps a stent placement to try to optimize renal function. And typically I will try to get a 24-hour urine creatinine clearance to really be sure a patient is cisplatin-ineligible before making these decisions. As I’m sure we can all comment on, cisplatin-eligible patients have a whole other realm of therapy available to them. We use dose-dense MVAC [methotrexate, vinblastine sulfate, doxorubicin, cisplatin]. There’s also gemcitabine and cisplatin, and I think for the first-line treatment those are still standard of care for patients who can get platinum, cisplatin.

Arlene O. Siefker-Radtke, MD:Gordon puts a stent in. Creatinine clearance goes up to 47 [mL/minute], and we get Betsy’s 24-hour urine, and she feels confident it’s at least 47. Would you give split dose cisplatin or would you give carboplatin?

Elizabeth R. Plimack, MD:So 47 mL/minute is a tricky number. I wish you had said 50. That’s sort of our threshold, especially in a 24-hour urine, which you typically get more points with in the 24-hour urine test than you would with the estimated. So assuming this patient, Cockcroft-Gault formula, is 42, let’s say we get him to 50, I would definitely give platinum with split dose. And there are different ways to give the split dosing. NCCN [National Comprehensive Cancer Network] Guidelines has some chemotherapy order sets and recommendations around that. But that really allows you to give the platinum over 2 days, the same amount of hydration really to protect the kidneys, and we’ve had success with that. That would be my approach in that setting.

Arlene O. Siefker-Radtke, MD:OK. It is something we’ve done at MD Anderson Cancer Center. We will use split dose cisplatin down to a GFR [glomerular filtration rate] of 40 mL/minute. And again, you have to be selective in your patients. You can’t really give this in a patient with a diabetic nephropathy. Those are the patients whose kidney functions won’t handle it. If the kidney is obstructed, you have to unobstruct the kidney with either a stent, or a nephrostomy tube is my preference because it does seem to do a better job of improving kidney function. But we’ve seen some data from IMvigor130 about chemotherapy plus immunotherapy [I/O]. And will this be a game-changer? Will it start changing people’s approaches who prefer cisplatin at split doses over carboplatin? I’d be curious as to what you think about the chemotherapy plus I/O group of patients. And we’ll start with Betsy.

Elizabeth R. Plimack, MD:The IMvigor130 trial is a randomized phase III trial conducted internationally. There were 3 arms. Arm A was atezolizumab plus chemotherapy. The chemotherapy was gemcitabine plus either carboplatin or cisplatin. Arm B was ATEZO [atezolizumab] monotherapy, and arm C was placebo plus that same chemotherapy. One important piece in understanding these data are that cisplatin-eligible and ineligible patients were enrolled, and the decision of whether to give CARBO [carboplatin] or cisplatin was a separate differentiator. So patients who were eligible for cisplatin didn’t always get it. Sometimes they were given carboplatin. The reverse was generally not true. It’s important in interpreting these data.

The baseline characteristics are typical of the patient population we see, and the end point for progression-free survival in the intent-to-treat arm comparing arm A, which is atezolizumab plus the chemotherapy, versus arm C, which is placebo plus chemotherapy, showed a progression-free survival advantage of 8.2 months for the atezolizumab plus chemotherapy arm versus 6.3 months for the chemotherapy alone arm. There are overall survival data as well, and comparisons were made between all of the different arms, which we can talk about later. But I think these data suggest that perhaps we should be thinking about combining atezolizumab with chemotherapy in the frontline setting. The key questions I think that these data raise but don’t answer are what to do with a cisplatin-eligible patient versus cisplatin-ineligible patient.

We know carboplatin and cisplatin aren’t the same. And so when we’re adding atezolizumab, does that same benefit benefit the cisplatin-eligible group as the cisplatin-ineligible group? Or is it limited to one or the other? They did do a forest plot, which suggests that more of the benefit is seen in the patients who get cisplatin, again different from eligible or ineligible. But very interesting data; in my mind, don’t quite change practice for this particular patient in your clinic right now.

Arlene O. Siefker-Radtke, MD:Looking at the trial, where I think they said upward of nearly half of patients who were cisplatin-eligible actually got carboplatin, or it was a relatively high number. How confident do you feel in that forest plot? Do you think there could be a little bit of cherry picking going on in the cisplatin arm, or do you think we’re seeing a real difference in the cisplatin, I/O versus carboplatin? I guess we haven’t seen much, but what do you think?

Tian Zhang, MD:Arlene, I think when we’re looking at this data, 52%, as you mentioned, of the cisplatin-eligible patients actually got carboplatin, would those patients have done better overall, would their progression-free survival been longer had they received cisplatin? That’s a key question that I think we’re all emphasizing. What I like about the data though is that in that atezolizumab plus chemotherapy arm…13% of patients actually achieved a complete response, which is very nice to see. I think what we’re seeing is an early rescue of stabilizing disease with the chemotherapy effect. And then as we get on into the later parts of the study where atezolizumab is given as maintenance, we’re seeing that durable effect of treatment.

And so perhaps the combination in a subset of patients may be beneficial to rescue these early progressors, people who won’t get on to a second-line therapy if they fail first-line. But in a disease setting where unfortunately we don’t have a lot of refractory lines of treatment options for these patients, I’m hesitant to use all of my active agents up front in a combination like this unless we see overall survival benefit later on.

Arlene O. Siefker-Radtke, MD:It sounds like you’re waiting until you see an overall survival benefit before you change your clinical practice. How about you, Betsy?

Elizabeth R. Plimack, MD:I absolutely agree. Again, if you’re using 2 agents up front that you would normally use sequentially, there has to be a signal of overall survival, otherwise it’s better to use them sequentially. It’s cost-effective; in terms of adverse effects, it kind of condenses those, so I agree.

Arlene O. Siefker-Radtke, MD:Do you think there could be any reason why giving chemotherapy with immunotherapy would be less beneficial? Do you think that chemotherapy could be impacting the immune response, and what we’re actually seeing as the impact of maintenance treatment with immunotherapy?

Elizabeth R. Plimack, MD:I’m really glad you asked that question. I think it’s hard for us to ever think that more would be not as good as a single agent. It’s not logical. But I think to be scientific we have to examine that possibility, and again that’s where the control arms are really important for this as we take a look. Part of the issue with control arms is you can’t always mandate a specific sequence, which is really our question. Is combination versus sequential meaningful? There were some data on subsequent therapies in IMvigor130 that were collected. Hopefully we’ll see that summarized.

Transcript edited for clarity.