Moshe C. Ornstein, MD, MA

 Case-Based Peer Perspective virtual event, Amir Mortazavi, MD, an associate Professor—Clinical, at The Ohio State University Comprehensive Cancer Center—Arthur G. James Cancer Hospital and Richard J. Solove Research Institute in Columbus, OH discussed available treatment regimen for a 

patient with clear cell renal cell carcinoma (RCC) and sarcomatoid features

Mortazavi is joined by Moshe Ornstein, MD, MA, a Genitourinary medical oncologist at the Taussig Cancer Institute of Cleveland Clinic in Cleveland, OH in the discussion.

: How do you assess patient risk, and does the risk status influence your treatment decision?

: We may not document [the risk status], but I think it’s important when that’s truly involved in your decision-making for different reasons.

There is a software available [to assess International Metastatic RCC Database Consortium (IMDC) risk that’s] basically calculation, and then you plug in [the risk factors]. Obviously, when there [are] none, then it’s good risk. If it’s 1 or 2 [factors], that’s intermediate risk. Three or more [factors] make it poor risk, which helps us decide, at least based on the evidence, which treatment to pick. But there are always other factors playing into that decision-making.

When a patient comes in with advanced disease, like in this case, what additional work-up would you do for the patient?

: Technically, if you go by the guidelines—guidelines are made by us, of course, mostly based on the data—the bone scan is indicated if you’re suspicious for bone mass either [due to] hypercalcemia, elevated alkaline phosphatase with your bone pain, or MRI or CT findings suggestive of bone mass. Then brain MRI is also indicative if there are symptoms, and that’s in most guidelines. Personally, if I have somebody with poor- risk bulky disease who is very symptomatic, before I start treatment, I do a brain MRI as long as the insurance doesn’t push back. I find you can handle many lower-volume asymptomatic brain masses with stereotactic brain radiotherapy, such as gamma knife, easily enough.

ORNSTEIN: I do the same. In general, I’ll for sure get brain imaging, and I get bone imaging if it’s clinically indicated.

What are your thoughts on the treatment options given in this poll?

: So if you look at the NCCN [National Comprehensive Cancer Network] guidelines for poor to intermediate, all 3 choices are acceptable options.

 Ipilimumab [Yervoy]/nivolumab [Opdivo] and axitinib [Inlyta]/pembrolizumab [Keytruda] have category 1 [recommendations] based on the data, and cabozantinib [Cabometyx] is an option, especially in a patient for whom you would consent to using immunotherapy.

ORNSTEIN: I’m always interested to hear how people make these decisions because, personally, I struggle with them. Part of the reason is 10 to 15 years ago, the median survival for patients with kidney cancer, metastatic RCC specifically, was probably less than 2 years. You look at the numerous options or numerous positive trials that we’ve had recently, including pembrolizumab and axitinib, avelumab [Bavencio] and axitinib, ipilimumab and nivolumab, cabozantinib, and a couple new trials that have not yet led to approvals but for which we are starting to have some data, and what’s interesting in all these trials [is the] OS [overall survival]. Pembrolizumab and axitinib demonstrated a hazard ratio for death of 0.53, and that was a statistically significant value [P < .0001]. Axitinib and avelumab were not statistically significant in terms of survival [HR, 0.78; P =.14], but ipilimumab and nivolumab were [HR, 0.68; P < .001].2 Cabozantinib versus sunitinib [Sutent] is not yet reported in terms of the P value, but the HR was 0.8, and the confidence interval crossed 1.

The nivolumab/cabozantinib is not [yet] approved. But all these other agents are approved in the frontline setting. And I think part of the discussion should revolve around this: When these agents are approved, how do we decide between giving somebody an immuno-oncology [IO]/tyrosine kinase inhibitor [TKI]–based regimen versus an IO-IO regimen versus monotherapy with a TKI?

The objective response rates [ORRs] seem to be the highest at 59% for pembrolizumab and axitinib and definitely higher in the immunotherapy-based regimens as opposed to with the single-agent TKIs. You’ll see different ranges of median follow-ups when the data were first reported. But there are a variety of options— the 2 most common ones being axitinib and pembrolizumab, and nivolumab and ipilimumab.

I normally give axitinib and pembrolizumab. Although for patients with sarcomatoid features, I have a little bit more of a debate.

What trial led to the approval of pembrolizumab/ axitinib as one of these recommended frontline treatment options?

ORNSTEIN: The KEYNOTE-426 trial [NCT02853331] was a phase 3 randomized trial that included patients of all IMDC risk groups. They were treatment naïve and were randomized to the combination of axitinib and pembrolizumab versus sunitinib, which at the time was really the standard of care in the frontline setting.

 Importantly, the dual primary end points were OS and progression-free survival [PFS] in the intent-to-treat [ITT] population. In some of the other trials, it was in the PD-L1–positive population or just in intermediate and poor risk. But KEYNOTE-426 was unique in the ITT populations across the board, all risk groups, and looked at OS and PFS, not based on PD-L1 staining, just all-comers, so fairly representative of what we see in the clinic.

The baseline characteristics [are] again representative of what we see in clinic, maybe a slightly higher percentage of favorable- risk patients. In clinic, we probably see 15% to 20% favorable-risk patients. And in this trial, in both arms, it was approximately 31%. The rest were intermediate and poor risk. The breakdown of PD-L1 based on the combined positive score [was about 60% with PD-L1 positivity]. And most of the patients in both arms have prior nephrectomies. We’ll highlight the sarcomatoid features in 17% or 18% of the patients essentially in both arms, and that’ll play a role as we look at some of the data.

Updated data at about 23-month minimum follow-up [show] the OS in the ITT population [median not reached with pembrolizumab/axitinib vs 35.7 months with sunitinib]—this was the primary end point—[had] an HR of 0.68 and a P value of less than .001. I think what’s important to take a look at, and this is one way I describe treatment decisions to patients, is looking at landmark OS. In other words, being able to tell patients, “Yes, with this combination in the trial and the attempt to treat all-comers, 90% of patients were alive at 12 months; 74% of patients were alive at 24 months.” It also helps me as I decide between treatment options, given we have so many these days. I’m getting a sense of those landmark analyses and seeing which regimens are potentially superior based on that.

The axitinib and pembrolizumab curves separated fairly early and seemed to be leveling off at the 33- to 36-month mark, demonstrating the promise of immunotherapy in the setting of having sustained responses over the course of time.

The other coprimary end point was the PFS in the ITT population [15.4 months with pembrolizumab/axitinib vs 11.1 months with sunitinib]—and again, significant values with an HR of 0.71 favoring axitinib and pembrolizumab. It’s a 24-month mark—in other words, at 2 years on therapy, 38% of patients had not yet had progression.

We don’t like comparing across trials, but we have to do that if we’re going to get any sense of how we decide on one regimen versus another. So if you look at the ORR for the patients on pembrolizumab/axitinib versus those patients on sunitinib, you see 60.2% for pembrolizumab/axitinib and 39.9% for sunitinib. I want to highlight the complete response [CR] rate [of 8.8% vs 3.0%, respectively]. The medium duration of response with this follow-up was 23.5 months [vs 15.9 months].

One concern you have when you see a patient with newly diagnosed [disease] is you want to get benefit ideally added to that initial regimen. So you have your 2 features favoring pembrolizumab/axitinib, and one of them is the response rate where 60% of patients responded in the trial. The other thing I want to draw your attention to is the best response of disease progression in only 11.3% of the patients. In terms of saying with confidence that you’re going to give patients a regimen and [that] 90% or so, at the very least, won’t have tumor progression at the time of their first scans, that’s an important thing to be able to tell patients.

We discussed a lot about the IMDC risk groups. So taking a deeper [look], because the patient in this case has intermediate/poor-risk metastatic RCC, at the OS Kaplan-Meier curve [by intermediate/poor risk]—that landmark OS at 24 months of 69% still being alive in the pembrolizumab/axitinib arm. If you also look at the separation between the curves and what at least appears to be a leveling off or the tail of the curve, which if it is at that mark around 33 to 36 months, that would be impressive. There was a separation in the intermediate- and poor-risk group for PFS, as well.

Would this treatment choice be a good option for this patient and why?

So the reason the sarcomatoid feature was thrown into the question stem was just to reinforce the idea that although sarcomatoid features in metastatic RCC are historically extremely difficult to treat and carry with them a very poor prognosis, it seems that with all regimens that are IO based in the 

, the responses have been incredibly impressive. Looking at the responses for patients on pembrolizumab/ axitinib if they had sarcomatoid features, the response rate was 59%, which was close to double the sunitinib arm.

 I’m telling you that although, historically, perhaps these patients would have done worse compared to the patients who did not have sarcomatoid features, using axitinib and pembrolizumab is certainly reasonable in patients with sarcomatoid features. In the waterfall plot [for patients with sarcomatoid features], looking at the overwhelming majority of patients, almost all of them had at least some tumor burden shrinkage.

The IO-based regimens for sarcomatoid features work, and you can see that in looking at the PFS for patients with sarcomatoid features treated with axitinib and pembrolizumab [not reached vs 8.4 months with sunitinib; HR, 0.54].

Which factors help you choose one regimen over the other?

: For me, when I see a patient who has a heavy disease burden—maybe has some symptoms, has some fevers, chills, is just systemically sick from the disease—the question I might ask is not necessarily how the patient is going to do in the long run but maybe more about which one of these combinations is most likely to give me a response right now. I think, for that patient, I would at least consider an IO-based regimen but with a TKI because the response rates are higher in those patients. The CR, maybe for a patient who needs a response right now, might be a bit less important. You might counter that the sarcomatoid features lead you to give full IO.

: It’s going to come down to each individual. So we’re not even going to talk about other comorbidities, how concerned you are about the immunotherapy part of adverse effects [AEs], or TKI AEs, and if somebody has some underlying autoimmune [disease], do you want to go with the pembrolizumab/axitinib? All [of these things] play a factor. What matters to the patient at the time of treatment is to get a response because the patient is very symptomatic; the tumors are getting to wider organs or going to drop the lung by blocking airway or encasing great vessels. All those play a factor like, “Do I need a response right now, or am I thinking more down the road?” I do the exact same approach. For that reason, I favor IO-based therapy. For the sarcomatoid features, we can talk within nivolumab/ipilimumab versus pembrolizumab/ axitinib. Just for that I think they both have a close response rate and also CRs, maybe nivolumab/ipilimumab a little more CR again, not to really do a cross comparison. That’s my approach to the treatment of frontline metastatic RCC.

What is the toxicity profile like for the pembrolizumab/ axitinib regimen compared with that of sunitinib?

: This is [similar to] what we already know about the TKIs. On the sunitinib side, you can see the diarrhea, hypertension, a lot of the on-target [effects of] VEGF and hypothyroidism, the hand-foot syndrome, and fatigue. [With] axitinib, we see some dysphonia, which is known to be more with axitinib, a little more hypertension, some higher-grade hypertension, but all along not hugely different. Typical incidences that we were expecting.

When it comes to more of the AEs of interest, autoimmune is going to obviously be more on the pembrolizumab/axitinib side. Under sunitinib, we still have hypothyroidism and hyperthyroidism to some degree. The incidence of immune-mediated AEs was higher with pembrolizumab/axitinib but not drastically [different from] what we know about the single-agent immune checkpoint inhibitors.

: The CheckMate 214 study [NCT02231749] was in a similar setting as the other study—that first-line patient with treatment-naïve metastatic clear cell RCC, stratification by IMDC and the region of enrollment. The patients were randomized 1:1 to nivolumab and ipilimumab versus sunitinib. They received standard-dose sunitinib, and they also had the dual primary end point of OS and PFS. Treatment went until disease progression or toxicity.

For the intermediate/poor-risk patient population at a 42-month follow-up, [there was a] 52% OS rate for the nivolumab/ipilimumab [arm] compared with 39% for sunitinib. And similar to the other studies, you get early separation of the curves, and it’s sustained with the tail in favor of nivolumab/ipilimumab. At 42 months, it has a ratio of 0.66 with a very strong P value.

As for PFS, at 42 months we have a 35% [PFS rate for nivolumab/ipilimumab] versus 19% [for sunitinib]; again, a strong P value [and HR] of 0.76, both primary end points favoring nivolumab/ipilimumab.

For the ORR, we see 42% versus 26%, so nivolumab/ipilimumab again superior, and we have a 10% [CR rate]. We’re not cross comparing, but higher than other studies for the CRs; the overall ORR altogether was lower than pembrolizumab/axitinib, but it’s still significantly higher than sunitinib.

As for the treatment-related AEs, [they were] very similar. [With] sunitinib, you see the TKI-AEs of hand-foot [syndrome] and gastrointestinal—diarrhea, nausea, vomiting—hypertension, and then cytopenia is more common. [With] nivolumab/ipilimumab, you have diarrhea and fatigue and skin toxicity with a different mechanism, itching, and then less of the electrolyte and cytopenias.

How did patients with sarcomatoid features fare in the CheckMate 214 trial?

: The nivolumab/ipilimumab trial had 139 patients who had sarcomatoid [features], and the axitinib/pembrolizumab trial had less. They had 51. Another trial that was reported had close to 70. So certainly the bulk was seen in the nivolumab/ ipilimumab trial. The response rates were slightly higher, and the CR rates were slightly higher as well, for patients with sarcomatoid [features] in the nivolumab/ipilimumab trial. In terms of the degree of tumor burden reduction, that was higher with axitinib and pembrolizumab, and maybe [that’s] not surprising because there’s a TKI in there.

I think that’s why most people who are polled would probably say that you can pick either one. You want to factor in how acutely a patient’s going to need a response up front. And you might want to have a TKI on board for that. But I think they’re fairly interchangeable. The HRs for PFS are nearly identical, but the HR for survival in sarcomatoid patients treated with nivolumab/ipilimumab was lower than it was for patients with axitinib and pembrolizumab.

How do you manage and mitigate AEs in patients receiving an IO-TKI combination?

: [Compared with] a lot of oncologists, I don’t treat a lot of malignancies. But for oncologists that give combination immunotherapy or a lot of immunotherapy for other malignancies, they might say that they’re extremely comfortable managing any -itis that shows up because you throw steroids at it, you taper the steroids, you figure out if you’re going to resume, but once a TKI is added to that, it can be a bit more complicated.

One of the reasons that I like this combination of axitinib and pembrolizumab is that the half-life for axitinib is extremely short; it’s about 4.5 hours. We know that there are overlapping toxicities; for instance, diarrhea. But if a patient is having a toxicity and you’re not sure if it’s coming from the checkpoint inhibitor or from the axitinib because the half-life is so short, if a patient stops axitinib for a couple of days and the AE goes away, then it’s probably from the axitinib. If the AE is still there, then you have to start considering that it might be coming from the immunotherapy.

As axitinib/pembrolizumab is approved, as cabozantinib/ nivolumab might be potentially coming down the pipeline, [as well as] lapatinib [Tykerb] and pembrolizumab—as these combined IO-TKI regimens are making their way into clinic, [it’s important to have] a practical approach to tell you if there’s a toxicity. You just stop the axitinib, and within 2 maybe 3 days, the AEs should be totally gone or mostly gone because of that extremely short half-life. That’ll tell you if this is a TKI-based toxicity or an IO-induced toxicity.

What are some of the management strategies for common AEs from the pembrolizumab/axitinib?

: For grade 1 diarrhea, you can just observe the patient, manage them symptomatically. Once it hits a grade 2, holding the axitinib for a couple of days, even just 24 to 48 hours should be sufficient. If they improve, then we can continue giving this patient axitinib. We might have to lower the dose [and/or] we might have to give them a couple of days of a break every few weeks because we know these to be on-target toxicities from axitinib. But if after a couple of days, there’s no improvement, well, then the culprit is most likely the immunotherapy, and then we would go down that pathway of treating patients with corticosteroids and deciding on other immunosuppressants, depending on response.

The same algorithm is going to show up essentially for all toxicities or for many of the toxicities, at least with the combination of something like axitinib and pembrolizumab.

What are your thoughts on the patient case at this point? How would you manage the patient now?

: The key is that the patient wasn’t having abdominal pain, cramping, or fevers, or things that usually come with colitis. And the assumption here was that it’s likely related to the axitinib.

: The management for this patient up front was basically to continue therapy and see if they could get control over the diarrhea. And once that didn’t happen, holding therapy for 3 days did resolve the diarrhea. Again, with the emphasis here being that we have to figure out, especially up front—this is 12 days into treatment, and you don’t want to give up on a regimen that has an OS benefit—how we’re going to manage this patient with what will, hopefully, be a chronic condition.

: For those who would hold and start antidiarrheals, I’m not saying that’s wrong, but that would not be what I would do. The reason is, when I hold, I want to see how the patient recovers. This was not a high-grade diarrhea that you have to slow down the output of the patient in just a grade 2 diarrhea. I prefer to hold and observe without any other intervention because that will help me understand better how quickly the patient recovers from this and clear the drug.

: I agree. I think you want to get as much information as possible. And again, for a [twice a day] drug, if you hold for 2, 3 days over the course of a month, they’re still getting the overwhelming majority.

I think holding the medication for a few days will give you more information. Then if you were to restart, you could consider restarting at the same dose with antidiarrheal. Or you could consider lowering the dose and seeing how the patient does.

I think it’s a matter of trying to tease out what you and your team are comfortable with. These patients are likely to be on therapy for a long time. You might have to do this a couple of times.

: The other thing I want to add is that you can just also pay attention to what else is happening to the patient. In my opinion, there’s a difference between the diarrhea from axitinib versus having a whole series of other issues—palmoplantar, and leukocytes, cytopenia, or other organ dysfunction. When I see all that, that’s usually when I know that I need dose reduction versus, in this case, when I would be quite comfortable with resuming the same dose.

What would your next step be in this patient, considering their case so far?

: Assuming the patient only had diarrhea, I would comfortably restart axitinib, but I wouldn’t start at the same dose. Some of these patients behave differently. You may see a rechallenging cause the same thing with the same pattern early on. Sometimes it happens later. I think there are other factors playing [a part here]. When I ask my patient what else they eat, you find a pattern—like when the patient is on these TKIs and they eat such and such, that triggers the diarrhea. So you can sometimes find those patterns and [then have] the patient just avoid [those foods]. So I would be comfortable to do the same dose for this patient.

The instruction this time is going to be “Just call us immediately,” especially if the patient says, “Yes, I waited 2 to 3 days.” And that’s a common pattern. So now they’re more educated on this issue and want to get us involved right away.

: So now the patient has a lot more symptoms. The diarrhea is worse, and you’re worried about more toxicity.

You want to do more work-up on a patient who’s quite symptomatic. So you’re going to hold a drug again and then schedule [the patient] to come for an office visit with blood work. You check the complete blood count, liver enzymes, chemistry, and you look for hepatitis and you look for nephritis. Then, for diarrhea, this time because it’s more severe; you want to rule out other causes and check parasites, especially [Clostridioides difficile] if they’ve been on antibiotics; they basically are going to need IV [intravenous] hydration. Because the patient was hypertensive, they hold the blood pressure medication.

Then the patient got supportive care, went home, and then called the next day and reported no diarrhea, feeling better, appetite is good, blood pressure is normal now, and all these stool tests are negative.

Do you agree with the approach for this patient up to this point?

: That’s a good thing because my approach is I stay full force until I get into the first set of scans. I’ll obviously do it safely, when I see the symptoms. I think we’re going to get into more of an overlapping toxicity: Now the patient has had a nice response to therapy but has elevated creatinine, whereas the creatinine may have been normal previously.

: I do the same. Again, I tried to be aggressive with symptomatic management, at least initially. If a patient is going to get to grade 2, grade 3, and we have to reduce, then that’s OK. The message I try to give to patients and other clinicians is just not to give up on the medication before we can see if it’s working. We can always lower the dose; we can always add some more antidiarrheals, assuming the patient is not dehydrated, to try to optimize management of the symptoms.

: I think the poor man’s test for nephritis is just to give steroids. If you give steroids and it gets better, well, you probably have your answer there. Sometimes I know the nephrologists get upset because they like to have a biopsy if possible; we get worried because the patient’s already down to 1 kidney, so we like to start steroids sooner. But I think the key is to hold both.

Again, if you have an improvement within a couple of days, it’s the [axitinib], and it’s fine. If you don’t, you probably want to start steroids sooner rather than later, unless you have a sameday nephrology appointment.

What would you do as the next steps for thispatient’s care?

 You have to figure out whether you’re going to restart the axitinib. So, now you have a scenario where the renal function is improving [and] you have a presumed nephritis.

I would have no issues restarting the axitinib because we have our answer as to what caused this AKI [acute kidney injury]. And then again, I’d be holding pembrolizumab until there’s complete renal function recovery. It’s important.

In summary, I think the key is that with axitinib, you have the benefit of having multiple doses. You have the benefit of a very short half-life. You have the benefit of being able to stop therapy, wait a couple of days, and see if there’s improvement to help you decide whether this is an IO-induced toxicity or a TKI-induced toxicity. So certainly leveraging that while managing patients can optimize their care.

1. NCCN. Clinical Practice Guidelines in Oncology. Kidney cancer, version 1.2021. Accessed January 19, 2021. https://bit.ly/363GcVf

2. Escudier B. Combination therapy as first-line treatment in metastatic renal-cell carci- noma. N Engl J Med. 2019;380(12):1176-1178. doi:10.1056/NEJMoa1900887

3. Choueiri TK, Hessel C, Halabi S, et al. Cabozantinib versus sunitinib as initial therapy for metastatic renal cell carcinoma of intermediate or poor risk (Alliance A031203 CABOSUN randomised trial): progression-free survival by independent review and overall survival update. Eur J Cancer. 2018;94:115-125. doi:10.1016/j. ejca.2018.02.012

4. Plimack ER, Rini BI, Stus V, et al. Pembrolizumab plus axitinib versus suni- tinib as first-line therapy for advanced renal cell carcinoma (RCC): updated analysis of KEYNOTE-426. J Clin Oncol. 2020;38(suppl 15):5001. doi:10.1200/ JCO.2020.38.15_suppl.5001

5. Rini BI, Plimack ER, Stus V, et al. Pembrolizumab (pembro) plus axitinib (axi) versus sunitinib as first-line therapy for metastatic renal cell carcinoma (mRCC): outcomes in the combined IMDC intermediate/poor risk and sarcomatoid subgroups of the phase 3 KEYNOTE-426 study. J Clin Oncol. 2019;37(suppl 15):4500. doi:10.1200/ JCO.2019.37.15_suppl.4500

6. Tannir NM, McDermott DF, Escudier B, et al. Overall survival and indepen- dent review of response in CheckMate 214 with 42-month follow-up: first-line nivolumab + ipilimumab (N+I) versus sunitinib (S) in patients (pts) with advanced renal cell carcinoma (aRCC). J Clin Oncol. 2020;38(suppl 6):609. doi:10.1200/ JCO.2020.38.6_suppl.609

7. Grünwald V, Voss MH, Rini BI, et al. Axitinib plus immune checkpoint inhibitor: evidence- and expert-based consensus recommendation for treatment optimisa- tion and management of related adverse events. Br J Cancer. 2020;123(6):898-904. doi:10.1038/s41416-020-0949-9

Articles

During a Targeted Oncology Case-Based Peer Perspective virtual event, Amir Mortazavi, MD andMoshe Ornstein, MD, MA discussed available treatment regimen for a patient with clear cell renal cell carcinoma and sarcomatoid features.


Experts Compare Available Regimens for a Patient With Sarcomatoid Features

Cisplatin-based chemotherapy has been and remains the standard of care in advanced urothelial carcinoma (aUC), and can provide a cure for a small subset of these patients. The clinical challenge in aUC is that most patients who are treated with cisplatin-based chemotherapy will ultimately develop refractory aUC and succumb to their disease. There is also a large cohort of patients who have other comorbidities that deem them ineligible to receive these regimens. Fortunately, since May 2016, 5 checkpoint inhibitors (CPIs) have been approved for the treatment of aUC. The rapid approval of these agents presents the clinician with challenges navigating drug selection and sequence decisions. We present an overview of the data that resulted in the approval of these agents, and present our approach to the treatment of patients with aUC.

Although a substantial percentage of patients with localized urothelial carcinoma (UC) will be cured primarily with local disease control, outcomes for patients with advanced disease remain exceedingly poor. For decades, the standard of care (SOC) for advanced UC (aUC) has been cisplatin-based chemotherapy, which results in a median survival of only approximately 15 months.

Although 13% to 15% of patients with aUC who are treated with chemotherapy can be cured,

most patients ultimately die of their disease.

Recently, novel immunotherapy with checkpoint inhibitors (CPIs) that target programmed cell death-1 (PD- 1) and programmed death ligand-1 (PD-L1) on T cells and tumor cells (TCs), respectively, has revolutionized therapy for patients with aUC. Five CPIs (atezolizumab, pembrolizumab, avelumab, durvalumab, and nivolumab) are now approved by the FDA for patients with aUC with relapsed/refractory disease following platinumbased chemotherapy. Similarly, 2 of these CPIs (atezolizumab and pembrolizumab) have also gained approval for patients with aUC who are cisplatin-ineligible.

The approval of multiple agents with similar mechanisms of action presents multiple dilemmas for the physician. Most notably, initial drug choice and sequence of therapies in aUC is vague and demands clarification. In this review, we present an overview of the pivotal trials and data that have resulted in the approval of these agents, and present our algorithmic approach for the treatment of patients with aUC.

A detailed overview of the role and data for chemotherapy in aUC is beyond the scope of this article. However, a brief overview is critical because cisplatin- based chemotherapy remains a cornerstone of therapy for patients with aUC. The combination of methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) was the historic SOC regimen in aUC, on the basis of results from a phase III trial that demonstrated improved overall survival (OS) with MVAC compared with cisplatin monotherapy (12.5 vs 8.2 months).

Given the relative unfavorable toxicity profile of traditional MVAC, efforts were undertaken to develop less-toxic, cisplatin-based combination therapies for patients with aUC.

Multiple phase II trials demonstrated that the combination of gemcitabine plus cisplatin (GC) was better tolerated than and had similar clinical efficacy to MVAC.

As such, a phase III trial comparing GC to MVAC in patients with aUC was conducted, with a primary endpoint of demonstrating the superiority of GC to MVAC.

In this trial, 405 patients were randomized in 1:1 fashion to receive either GC or MVAC for a maximum of 6 cycles. The primary objective of OS was similar between the GC and MVAC cohorts (14.0 vs 15.2 months; HR, 1.09 [95% CI, 0.88-1.34]; P = .66), as were the 5-year survival rates (GC,3.0; MVAC, 15.3; P = .53).

Other efficacy endpoints such as objective response rate (ORR), time to progressive disease, and time-to-treatment failure were also similar between the 2 arms. The toxicity profile favored the GC cohort, in which there was less grade 3/4 neutropenia (71% vs 82%), less neutropenic fever (2% vs 14%), and a lower treatment-related mortality rate (1% vs 3%) compared with the classic MVAC cohort.

Although the trial was not designed to demonstrate noninferiority of GC compared with MVAC, many clinicians rely on these data to support the use of GC as a standard frontline therapy for aUC, given similar efficacy and less toxicity.

Other methods of improving toxicity of traditional GC and MVAC, in which the same drugs are given on slightly different schedules, have been thoroughly investigated and incorporated into clinical practice.12,13 Perhaps most notable was the introduction of the high-dose-intensity MVAC (HD-MVAC; also known as dose-dense MVAC [dd-MVAC]) regimen, in which patients are treated with MVAC every 2 weeks (standard dosage of each drug) with granulocyte colony-stimulating factor (G-CSF) support instead of the 4-week traditional (classic) MVAC regimen. In an international phase III trial comparing classic MVAC versus HD-MVAC in aUC, patients receiving HD-MVAC had less toxicity, fewer treatment delays, and were able to receive twice the number of cisplatin doses as the classic MVAC cohort, despite being on a 14-day treatment regimen.

In a long-term follow-up of this study, patients who received HD-MVAC had improved ORR, complete response (CR), and progression-free survival (PFS) compared with those who received classic MVAC.14 Thus, HD-MVAC is a generally accepted standard over the traditional MVAC in this setting.

Despite these dosage and schedule modifications, cisplatin-based chemotherapy remains relatively toxic. A primary challenge, therefore, is determining which patients should be given these chemotherapy regimens. A consensus panel developed guidelines to determine cisplatin eligibility. The consensus guidelines declare a patient as &ldquo;cisplatin-eligible&rdquo; if they meet the following criteria: ECOG PS (European Cooperative Oncology Group performance status) 0-1, creatinine clearance (CrCl) &ge;60 mL/min, no grade &ge;2 peripheral neuropathy, no New York Heart Association Class III/IV heart failure, and no significant hearing loss.

However, the optimal cutoff of adequate renal function is being discussed, with a number of experts administering cisplatin with CrCl &ge; 50 mL/min, often with &ldquo;split dose&rdquo; cisplatin (dose divided over 2 days) in borderline cases, while 24- hour urine can be used in such cases to calculate more accurately CrCl. Given the level I (based on randomized phase III trials) evidence of cisplatinbased chemotherapy in aUC and 5-year survival rates of approximately 15%, patients with aUC who meet the eligibility criteria for cisplatin should thus be treated with either GC or ddMVAC as SOC frontline therapy (

Postchemotherapy Checkpoint Inhibitors in Platinum-Refractory aUC

Historically, patients who progressed on or following platinum-based chemotherapy for aUC were treated with single-agent chemotherapy such as paclitaxel or docetaxel (or vinflunine in Europe).

However, median PFS and OS with these agents are only approximately 3 and 7 months, respectively, underscoring the need for alternative therapies. Recently, multiple CPIs were approved in platinum-refractory aUC, offering superior clinical options for these patients (Table 1).

The first CPI approved for platinum-refractory aUC was atezolizumab (antiPD-L1). IMvigor 210

was a single-arm phase II trial investigating the use of atezolizumab in patients with aUC. This trial comprised 2 cohorts; cisplatin-ineligible patients (cohort 1); and patients who progressed after platinum- based therapy (cohort 2). In cohort 2, a total of 310 patients were treated with atezolizumab at the standard dosage of 1200 mg every 3 weeks. The primary endpoint of ORR was 15% in all patients. Of note, patients who had increased PD-L1 expression of tumor-infiltrating immune cells (IC0: <1%; IC1: &ge;1% and <5%; IC2/3 &ge;5%) had higher response rates than those with lower IC expression. IC2/3 patients had ORR of 26%, and IC1/2/3 demonstrated ORR of 18%. Responses were also durable, with 84% of patients having an ongoing response at a median follow-up of 11.7 months. Interestingly, mutational load and The Cancer Genome Atlas (TCGA) luminal II subtype were predictive of higher response rate to treatment. Median OS was 7.9 months in all patients, but was significantly higher in the IC2/3 cohort (11.4 months).

Adverse events (AEs) are noted in Table 1.

Based on these data, on May 18, 2016, the FDA granted accelerated approval for atezolizumab in patients with platinum-refractory aUC. There are a number of interesting features of this approval. In addition to the approval of atezolizumab, the FDA concurrently approved the complementary diagnostic VENTANA PD-L1 (SP142) assay. Importantly, despite its approval as a complementary diagnostic, atezolizumab was approved regardless of PD-L1 expression because of objective and durable responses noted, even with low PD-L1 expression.6 It is also interesting to note that this approval was based on a single-arm trial in the context of historical control data (ie, historical response rates of chemotherapy in the platinum-refractory setting). This highlights the dire need for effective treatments in this setting.

Atezolizumab was granted accelerated approval while the phase III IMvigor 211 confirmatory study of atezolizumab versus chemotherapy was enrolling patients.

A press release by Roche on May 10, 2017, announced that the IMvigor 211 study did not meet its primary endpoint of OS compared with chemotherapy in the predefined patient subset with IC2/3 PDL1 expression.

Although the full data from this trial have not yet been published, atezolizumab retains its accelerated FDA approval for platinumrefractory aUC, and was recently also approved by the European Medicines Agency (EMA).

The KEYNOTE-045 trial was a phase III, international trial that randomized 542 patients with platinum-refractory disease to receive either pembrolizumab (antiPD-1) 200 mg every 3 weeks or chemotherapy per investigator choice of docetaxel, vinflunine, or paclitaxel.

Median OS (mOS) in the entire cohort, irrespective of PD-L1 expression, was significantly higher in pembrolizumab-treated patients compared with those treated with chemotherapy (10.3 vs 7.4 months; HR, 0.73 [95% CI, 0.59- 0.91]; P = .002). PD-L1 expression was investigated using a combined positive score (CPS), which calculated PD-L1 expression on TCs and ICs relative to the total number of TCs. In patients with a CPS of &ge;10, the mOS was superior in the pembrolizumabversus- chemotherapy cohorts (8.0 vs 5.2 months; HR, 0.57 [95% CI, 0.37-0.88]; P = .005).4 Treatmentrelated AEs were also less common in the pembrolizumab arm (Table 1). In an updated analysis, ORR was 21.1% for pembrolizumab versus 11.0% for chemotherapy- treated patients. The median duration of response (DoR) was not reached for pembrolizumab (1.6 - 20.7+) and was 4.4 months (1.4-20.3+) for the chemotherapy cohort.

Pembrolizumab thus received regular approval for patients with platinumrefractory aUC, and was also approved by the EMA. In addition to the clinical trials of pembrolizumab and atezolizumab, which have data regarding their phase III trial results either in press release and presented abstract (atezolizumab) or academically published (pembrolizumab) form, 3 other CPI agents are currently approved in platinum-refractory aUC. These agentsnivolumab (antiPD-1), avelumab (anti&ndash;PD-L1), and durvalumab (anti&ndash;PD-L1)&mdash;are all approved based on results from single-arm phase I/ II clinical trials

Choosing an Immunotherapy in Platinum-Refractory aUC

In the now-crowded platinum-refractory aUC setting, choosing between CPI options has not been clearly defined. Our approach is, therefore, not based on rigorous head-to-head comparisons of agents, but rather based on the current and emerging level of evidence supporting these agents. Nivolumab, durvalumab, avelumab, and atezolizumab are approved on the basis of phase I/II data. Moreover, the IMvigor phase III trial data for atezolizumab&nbsp;was reported as not having met its primary endpoint. However, pembrolizumab is approved on the basis of the results from a phase III clinical trial. Given the superiority of this level of evidence, it is reasonable to use pembrolizumab in platinum-refractory aUC. However, any of the above agents are appropriate based on FDA approval, and all 5 appear to have comparable efficacy and safety profiles, although any cross-comparisons between clinical trials have inherent limitations. Clinical trials remain a very critical part of new treatment evaluation and should always be considered.

Historically, patients who were not clinically fit to be treated with cisplatin were treated with other chemotherapy regimens (eg, carboplatin-based chemotherapy), which, although active, are significantly less effective than cisplatin.

More recently, however, that paradigm has shifted as CPIs have been incorporated into first-line therapy for cisplatin- ineligible patients with aUC, based on results of phase II, single-arm clinical trials (Figure; Table 2).

As previously mentioned, IMvigor 210 was a single-arm, 2-cohort phase II trial investigating the use of atezolizumab in patients with aUC.

In cohort 1 (cisplatin-ineligible patients), 119 patients were treated with atezolizumab 1200 mg every 3 weeks until progression. With a median follow-up of 17.2 months, the ORRs and CRs were 23% and 9%, respectively, with a median OS of 15.9 months (10.4-not reached [NR]). At a median follow-up of 17.2 months, 70% of responses were still ongoing.

Immune-mediated AEs occurred in 12% of patients. Interestingly, higher mutational load was associated with higher response rates.

These impressive efficacy and safety results in this setting led to accelerated FDA approval of atezolizumab as a frontline treatment option for cisplatin-ineligible patients with aUC, and atezolizumab also was recently approved by the EMA

KEYNOTE-052 was a single arm, phase II trial of pembrolizumab in the first-line setting for cisplatinineligible patients.

These 370 patients were treated with pembrolizumab 200 mg every 3 weeks for a maximum of 24 months or until progressive disease (PD), intolerable toxicity, or withdrawal of consent. With a median follow-up of 9.5 months, the primary endpoint of ORR was 29% (7% CR). The median time to response was 2 months (range, 1-9 months). Median DoR was not reached (12 months NR). Impressively, 82% of responses lasted more than 6 months.

Adverse events are highlighted in Table 2. Efficacy and safety appear similar, even in the subset of patients with poor PS and in the elderly (age &ge;75 years). Based on these data, the FDA granted accelerated approval to pembrolizumab in cisplatin-ineligible aUC, and it was also recently approved in Europe.

Choosing Therapy in Cisplatin-Ineligible Patients

Despite advances with CPIs, it is important to note that patients who are cisplatin-ineligible may still be treated with other chemotherapeutic agents, such as carboplatin-based regimens (Figure).

However, patients who are generally unfit for any chemotherapy due to medical comorbidities, poor performance status, or other reasons may receive CPI therapy as SOC. Similarly, genomic sequencing of circulating tumor DNA and tumor tissue can be used to guide therapy (Figure).

Clinical trials are an important option, as well, and should be considered in the appropriate setting.

With regard to CPI treatment options in cisplatinineligible and chemotherapy-ineligible patients with aUC, the efficacy of pembrolizumab and atezolizumab appears comparable, taking into account the limitations of cross-trial comparisons (Table 2). The level of evidence is also identical as both trials were single-arm, noncomparator, phase II trials. Therefore, both are appropriate agents in this clinical setting. An institution may implement a policy in which a specific CPI is on formulary based on other cancer types and/or treatment settings (eg, data as noted above regarding pembrolizumab in platinum-refractory aUC). Such institutional selection, available guidelines, insurance policies, frequency of administration, and costeffectiveness studies in the &ldquo;real world&rdquo; setting may ultimately guide therapeutic choice of CPI in different settings. Importantly, however, from an evidence-based approach, both atezolizumab and pembrolizumab are acceptable agents in the cisplatin-ineligible, frontline aUC setting.

Chemotherapy has been the cornerstone of aUC management and still has an established role, especially cisplatin-based chemotherapy in the first-line setting. A number of clinical trials in this setting are comparing chemotherapy to immunotherapy, combinations of chemotherapy and immunotherapy, or sequential strategies such as switching to maintenance immunotherapy after an initial response or stable disease to chemotherapy. The emergence of CPIs in both cisplatin-ineligible and platinumrefractory aUC has resulted in a much-needed improvement in clinical outcomes in a number of patients; however, more progress is urgently and clearly required. Pembrolizumab and atezolizumab are options in the cisplatin-ineligible setting, with pembrolizumab owning the highest level of evidence in platinum-refractory disease.

We acknowledge that all of the agents mentioned above are FDA approved and appropriate for clinical practice in the appropriate setting. Of note, there is a plethora of clinical trials as well as translational We acknowledge that all of the agents mentioned above are FDA approved and appropriate for clinical practice in the appropriate setting. Of note, there is a plethora of clinical trials as well as translational&nbsp;research investigating the molecular biology of UC (eg, germline and somatic mutations, gene expression profiling, mutational load, PD-L1 expression, other biomarkers).

Our hope and anticipation is that these research endeavors will ultimately drive treatment selection and further improve outcomes for patients with aUC.

von der Maase H, Hansen SW, Roberts JT, et al. Gemcitabine and cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in advanced or metastatic bladder cancer: results of a large, randomized, multinational, multicenter, phase III study.

von der Maase H, Sengelov L, Roberts JT, et al. Long-term survival results of a randomized trial comparing gemcitabine plus cisplatin, with methotrexate, vinblastine, doxorubicin, plus cisplatin in patients with bladder cancer.

Apolo AB, Infante JR, Balmanoukian A, et al. Avelumab, an anti-programmed death-ligand 1 antibody, in patients with refractory metastatic urothelial carcinoma: results from a multicenter, phase ib study.

2017;35(19):2117-2124. doi: 10.1200/JCO.2016.71.6795.

Bellmunt J, de Wit R, Vaughn DJ, et al. Pembrolizumab as secondline therapy for advanced urothelial carcinoma.

. 2017;376(11):1015-1026. doi: 10.1056/NEJMoa1613683.

Powles T, O&rsquo;Donnell PH, Massard C, et al. Efficacy and safety of durvalumab in locally advanced or metastatic urothelial carcinoma: updated results from a phase 1/2 open-label study.

2017;3(9):e172411. doi: 10.1001/jamaoncol.2017.2411.

Rosenberg JE, Hoffman-Censits J, Powles T, et al. Atezolizumab in patients with locally advanced and metastatic urothelial carcinoma who have progressed following treatment with platinum-based chemotherapy: a single-arm, multicentre, phase 2 trial.

. 2016;387(10031):1909- 1920. doi: 10.1016/S0140-6736(16)00561-4.

Sharma P, Retz M, Siefker-Radtke A, et al. Nivolumab in metastatic urothelial carcinoma after platinum therapy (CheckMate 275): a multicentre, single-arm, phase 2 trial.

. 2017;18(3):312-322. doi: 10.1016/S1470-2045(17)30065-7.

Balar AV, Galsky MD, Rosenberg JE, et al. Atezolizumab as first-line treatment in cisplatin-ineligible patients with locally advanced and metastatic urothelial carcinoma: a single-arm, multicentre, phase 2 trial.

. 2017;389(10064):67-76. doi: 10.1016/S0140-6736(16)32455-2.

Loehrer PJ S, Einhorn LH, Elson PJ, et al. A randomized comparison of cisplatin alone or in combination with methotrexate, vinblastine, and doxorubicin in patients with metastatic urothelial carcinoma: a cooperative group study.

von der Maase H. Gemcitabine in locally advanced and/or metastatic bladder cancer.

von der Maase H, Andersen L, Crino L, Weinknecht S, Dogliotti L. Weekly gemcitabine and cisplatin combination therapy in patients with transitional cell carcinoma of the urothelium: a phase II clinical trial.

Sternberg CN, de Mulder PH, Schornagel JH, et al; European Organization for Research and Treatment of Cancer Genitourinary Tract Cancer Cooperative Group. Randomized phase III trial of high-dose-intensity methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) chemotherapy and recombinant human granulocyte colony-stimulating factor versus classic MVAC in advanced urothelial tract tumors: European Organization for Research and Treatment of Cancer protocol no. 30924.

Als AB, Sengelov L, Von Der Maase H. Gemcitabine and cisplatin in locally advanced and metastatic bladder cancer; 3- or 4-week schedule?

Sternberg CN, de Mulder P, Schornagel JH, et al; EORTC Genito-Urinary Cancer Group. Seven year update of an EORTC phase III trial of high-dose intensity M-VAC chemotherapy and G-CSF versus classic M-VAC in advanced urothelial tract tumours.

Galsky MD, Hahn NM, Rosenberg J, et al. Treatment of patients with metastatic urothelial cancer &ldquo;unfit&rdquo; for cisplatin-based chemotherapy.

2011;29(17):2432-2438. doi: 10.1200/JCO.2011.34.8433.

Raggi D, Miceli R, Sonpavde G, et al. Second-line single-agent versus doublet chemotherapy as salvage therapy for metastatic urothelial cancer: a systematic review and meta-analysis.

. 2016;27(1):49-61. doi: 10.1093/annonc/mdv509.

A study of atezolizumab compared with chemotherapy in participants with locally advanced or metastatic urothelial bladder cancer [IMvigor211]. clinicaltrials.gov/ct2/show/NCT02302807. Updated December 13, 2017. Accessed December 10, 2017.

Roche provides update on phase III study of TECENTRIQ (atezolizumab) in people with previously treated advanced bladder cancer [press release]. www.roche.com/media/store/releases/med-cor-2017-05-10.htm. Basel, Switzerland: Roche; May 10, 2017.

Bajorin DF, DeWit R, Vaughn DJ, et al: Planned survival analysis from KEYNOTE-045: phase 3, open-label study of pembrolizumab (pembro) versus paclitaxel, docetaxel, or vinflunine in recurrent, advanced urothelial cancer (UC).

Massard C, Gordon MS, Sharma S, et al. Safety and efficacy of durvalumab(MEDI4736), an anti-programmed cell death ligand-1 immune checkpoint inhibitor, in patients with advanced urothelial bladder cancer.

. 2016;34(26):3119-3125. doi: 10.1200/JCO.2016.67.9761.

Sharma P, Callahan MK, Bono P, et al. Nivolumab monotherapy in recurrent metastatic urothelial carcinoma (CheckMate 032): a multicentre, open-label, two-stage, multi-arm, phase 1/2 trial.

2016; 17(11):1590-1598. doi: 10.1016/S1470-2045(16)30496-X.

Galsky MD, Chen GJ, Oh WK, et al. Comparative effectiveness of cisplatin-based and carboplatin-based chemotherapy for treatment of advanced urothelial carcinoma

. 2012;23(2):406-410. doi: 10.1093/annonc/mdr156.

De Santis M, Bellmunt J, Mead G, et al. Randomized phase II/III trial assessing gemcitabine/carboplatin and methotrexate/carboplatin/vinblastine in patients with advanced urothelial cancer who are unfit for cisplatin-based chemotherapy: EORTC study 30986.

. 2012;30(2):191-9. doi: 10.1200/JCO.2011.37.3571.

O&rsquo;Donnell PH, Grivas P, Balar AV, et al: Biomarker findings and mature clinical results from KEYNOTE-052: first-line pembrolizumab (pembro) in cisplatin-ineligible advanced urothelial cancer (UC).

Barata PC, Koshkin VS, Funchain P, et al. Next-generation sequencing (NGS) of cell-free circulating tumor DNA and tumor tissue in patients with advanced urothelial cancer: a pilot assessment of concordance.

. 2017;28(10):2458-2463. doi: 10.1093/annonc/mdx405.

Cisplatin-based chemotherapy has been and remains the standard of care in advanced urothelial carcinoma, and can provide a cure for a small subset of these patients.

Moshe C. Ornstein, MD, MA | Authors

Articles

Experts Compare Available Regimens for a Patient With Sarcomatoid Features

Selecting and Sequencing Therapy in Advanced Urothelial Carcinoma: Navigating a Dynamic Landscape