ASCO 2021: Optimizing Urothelial Cancer Management From Organ-Confined to Metastatic Disease: A Multidisciplinary Approach – Medical Oncologist Perspective

( This session was a multidisciplinary case discussion focusing on multiple topics including (1) treatment of BCG unresponsive non-muscle invasive bladder cancer, (2) treatment of muscle-invasive urothelial carcinoma in cisplatin eligible and ineligible patients, (3) the implications of pathologic response to neoadjuvant chemotherapy, (4) the role of adjuvant therapy, (5) the management of metastatic urothelial carcinoma and the role of chemotherapy, immune therapies, FGFR inhibition, and antibody-drug conjugates, and finally (5) palliative therapy.

Case: 62F presents with gross hematuria and undergoes cystoscopy.


She is diagnosed with carcinoma in situ (CIS) and receives intravesicular BCG for six treatments followed by maintenance therapy. After six months of maintenance therapy, her surveillance cystoscopy reveals recurrence carcinoma and situ and a new invasive T1 urothelial carcinoma.


Dr. Hahn discussed that this would be termed BCG-unresponsive non-muscle-invasive bladder cancer. The term BCG-unresponsive, which has evolved over time through workshops and FDA consultations, is defined as recurrence of carcinoma in situ within 12 months of adequate BCG, or in a patient without CIS, if they had high-grade papillary disease recurrence within six months of adequate BCG. Adequate BCG is defined by the 5+2 rule, meaning they needed to have received at least 5 of the 6 planned induction treatments, then at least 2 of 3 maintenance therapies. We do not typically recommend additional BCG treatments for BCG-unresponsive patients.

The first question posed was “What treatment option would you proceed with?”. (A) Intravesical chemotherapy, (B) Intravenous pembrolizumab, (C) Radical Cystectomy, or (D) Other.

Dr. Hahn discussed some of the data behind intravenous pembrolizumab if the patient is not interested in cystectomy. This is an FDA-approved therapy for BCG-unresponsive patients with a component of CIS. Though this particular patient would have been eligible for KEYNOTE-57, which established this treatment as standard of care, it is important to note that in that trial the majority of patients were flat CIS alone, and only 1/3 had an associated high-grade papillary component. Based on data with advanced disease, where 15-20% of patients have disease durable control, we expect some patients to get long-term disease control in the non-muscle invasive patient as well. There are no clear biomarkers such as PD-L1 status that help identify patients more likely to respond. It is important to understand more deeply how the patient thinks about cystectomy because for pT1 disease it is the therapy that gives them the best chance for long-term durable disease control.

The patient is not interested in radical cystectomy or intravenous pembrolizumab due to travel requirements. She is willing to receive intravesical chemotherapy with gemcitabine and docetaxel. After intravesical chemotherapy, surveillance cystoscopy performed 6 months later reveals invasive disease obscuring the bilateral ureteral orifices, resulting in kidney injury. CT imaging reveals bilateral hydronephrosis but no significant lymphadenopathy.

From a medical oncology perspective, relieving ureteral obstruction with bilateral nephrostomy tubes helps us understand quickly if the patient can recover enough renal function to receive cisplatin-based chemotherapy. Second, it helps avoid start/stops with cisplatin, because if the renal function is worsening on the therapy you don’t have to consider ureteral obstruction as a potential reason. However, we must also be sympathetic to the quality-of-life decrease associated with these external drains and managing them. I try to emphasize especially in the pre-operative setting, that the goal is to facilitate the best chance of cure and hopefully the tubes will be only in for 2-3 months.

Bilateral percutaneous nephrostomy tubes are recommended, and the patient wants to start treatment as soon as possible. Her ECOG PS is 1 and she has no other major comorbidities. What options do you consider? (A) Radical cystectomy, (B) Split-dose cisplatin-combination chemotherapy followed by cystectomy, (C) carboplatin-combination chemotherapy followed by cystectomy, (D) pembrolizumab followed by cystectomy, (E) Bladder preservation with cisplatin-based chemoradiation, (F) Bladder preservation with 5FU based chemoradiation.

Based on the patient’s kidney injury, she would be likely ineligible for cisplatin, and at 38 mL/min, split dose chemotherapy is also concerning. We now have published data from Matt Galsky in JCO that set some uniform definitions of cisplatin eligibility, including performance status (0 or 1) and creatinine clearance of around 60. Other factors are also important such as baseline neuropathy, ototoxicity, or heart failure. However, in a patient that could receive curative therapy, many practitioners would lower the creatinine clearance threshold for cisplatin eligibility. Many people are comfortable going down to 50 if using split-dose cisplatin regimens, and this can be pushed a little further if nephrostomy tubes are in place and we can do adequate hydration, but 38 is a bit too low. Carboplatin can certainly be given, but I am not as comfortable with this in a patient who is cystectomy eligible. The role of immunotherapy is interesting given data from PURE-01 (pembrolizumab single agent) and ABACUS (single-agent atezolizumab). Even with brief exposure of three cycles relative to three months with chemotherapy, there is a complete response rate of 20-30%, but this still lives really in the clinical trial realm given that we do not know the long-term implications of complete response to immunotherapy. Multiple trials are ongoing, such as KEYNOTE-905 looking at pembrolizumab versus enfortumab vedotin (anti-nectin4 antibody-drug conjugate) versus cystectomy in cisplatin-ineligible patients.

With bilateral nephrostomy tubes, the creatinine clearance improves to 61 ml/min and becomes cisplatin eligible. She prefers surgery but would be willing to receive neoadjuvant chemotherapy if recommended. Do you agree or disagree?

Dr. Hahn agreed with neoadjuvant chemotherapy, and as a general statement does not have a clear preference between regimens. Both dose-dense MVAC and cisplatin/gemcitabine are accepted regimens, and in retrospective studies, there was no difference in pathologic complete responses or overall survival. Furthermore, in SWOG-1314 (COXEN), there was no difference in pathologic complete response in a randomized phase 2 trial between these two regimens, with an approximate 30% complete response rate.

The patient undergoes neoadjuvant dd MVAC and then radical cystectomy with an ileal conduit. Pathology reveals ypT2N0 disease. What course of action would you recommend next? (A) Adjuvant carboplatin with gemcitabine, (B) Adjuvant taxane-combination chemotherapy regimen, (C) Adjuvant checkpoint inhibitor, (D) Observation

Current guidelines recommend observation as there are no prospective data proving an advantage for any adjuvant chemotherapy, just retrospective and potentially flawed data. There have been two trials that have read out in this context. The IMvigor010 study of 800 patients looking at the impact of adjuvant atezolizumab in high-risk patients. No disease-free survival benefit was seen with the atezo treated group, and there was no clear subgroup that benefitted from therapy. CheckMate-274 looking at a similar population of over 700 patients showed a significant disease-free survival advantage for adjuvant nivolumab (21 months with nivolumab versus 10.9 months with placebo), but no overall survival data is available yet. Finally, we have the AMBASSADOR trial of adjuvant pembrolizumab that has not read out yet. As of today, I am not offering adjuvant checkpoint inhibitor therapy, but based on regulatory agency approvals and further clinical trial data, this may change.

The patient undergoes observation. Six months later, surveillance CT imaging reveals the following image. Creatinine clearance is now 48 ml/min and ECOG performance status is 1. Would you perform PD-L1 IHC on archival tissue to aid in treatment decision-making?


For a patient with visceral liver metastases, I am not reaching towards immunotherapy first, and so PD-L1 IHC would not impact my treatment choice. If the patient is frail and not enthusiastic about carboplatin-based chemotherapy, then PD-L1 testing is useful especially if the label indication for immunotherapy requires it. At this point, given the quick relapse, this could be considered second-line therapy. While the responses to immunotherapy are not as good in liver metastases, patients with low-volume liver disease burden will have a chance of durable long-term disease control with immunotherapy.

PD-L1 testing is performed with the 22C3 antibody, and the CPS score returns at < 1. What course of action would you take next? (A) Gemcitabine with cisplatin, (B) Gemcitabine with carboplatin, (C) pembrolizumab, (D) gemcitabine with carboplatin plus atezolizumab

This is a tough question because the patient has relapse so soon after cisplatin-based chemotherapy, so would still be considering pembrolizumab therapy given the chance of durable response. Patients with a CPS score of less than 1 do have responses to immunotherapy. This is not a thrilling choice, but Dr. Hahn would favor pembrolizumab here.

Agree or Disagree: If the patient relapsed after 15 months instead of 6 months after neoadjuvant chemotherapy, that would change my systemic treatment selection.

Dr. Hahn would think more favorably about giving chemotherapy in this context, especially with the CPS score of less than 1. The chance of durable response in this patient is not high, but the presence of liver metastases is concerning for impending rapid expansion of systemic disease and so chemotherapy gives a better shot at a response. Dr. Hahn does not think it would be wrong to give immunotherapy given the chance of durability. Given that it is 12 months post neoadjuvant chemotherapy, there is an FDA label for immunotherapy if PD-L1 expression is high. However, he would favor chemotherapy more.

Agree or Disagree: I would obtain next-generation sequencing of the tumor at this time

Yes, given that an FDA-approved therapy for certain FGFR alteration-positive urothelial cancers, Dr. Hahn would sequence the tumor.

Foundation Medicine sequencing reveals a microsatellite stable tumor without hypermutation. There is no FGFR alteration or other actionable genomic alteration. The patient has a response to pembrolizumab for one year before liver metastases begin to grow and new lung and bone metastases are found on restaging imaging. What therapy would you proceed with next? (A) Taxane chemotherapy, (B) enfortumab vedotin, (C) Erdafitinib, (D) Sacituzumab govitecan

Dr. Hahn would favor enfortumab vedotin for two reasons: (1) presence of liver metastases where other therapies like taxanes have response rates of 10% but enfortumab has response rates in the 30-40% range, and (2) enfortumab has an overall survival advantage demonstrated compared to additional non-platinum-based chemotherapy in the randomized phase 3 trials. Because there is no proven overall survival advantage yet with Sacituzumab, though a promising strategy, he would go with enfortumab first.

The patient responds to enfortumab vedotin for 9 months before developing increasing back pain. ECOG PS is now 3. After obtaining a spine MRI, what will be your next course of action? (A) Hospice, (B) Radiation therapy, (C) Surgical decompression, (D) Palliative chemotherapy and bisphosphonates


The session ended with key take-home points, which are shown below.


Presented by: Noah M. Hahn, MD, Medical Oncologist at the Johns Hopkins University School of Medicine, Baltimore, MD

Written by: Alok Tewari, MD, Ph.D., Medical Oncologist at the Dana-Farber Cancer Institute, at the virtual 2021 American Society of Clinical Oncology Annual Meeting Congress (#ASCO21), June 4th-June 8th, 2021