Kamaneh Montazeri & Guru Sonpavde
Abstract
Introduction: Metastatic urothelial carcinoma (mUC) remains a fatal malignancy, despite the recent addition of immune checkpoint inhibitors (ICIs), an FGFR inhibitor and an antibody-drug conjugate (ADC) to the therapeutic armamentarium. The survival rates are particularly dismal after first-line treatment failure, entailing an urgent need for more effective therapies. Advances in understanding biomarkers and identifying targetable molecules have broadened the pathways under investigation in mUC. Areas covered: This review summarizes mUC salvage therapy options, including chemotherapy, ICI, and novel promising agents, including targeted therapies, ADCs, cytotoxic agents and vaccines. For the literature review, a PubMed search and relevant data presented at international conferences were used. Expert opinion: The approval of ICIs, FGFR inhibitor erdafitiniband ADC enfortumab vedotin in the salvage setting has transformed the mUC landscape. Yet there are additional promising agents currently under study. Toxicities are observed with ADCs and FGFR inhibitors, but appear manageable in most patients. The molecular heterogeneity and complex tumor biology are challenging barriers for progress in the therapy of mUC. Advances in molecular profiling, defining validated predictive markers, rational combinations of agents and therapeutically actionable targets will help develop personalized compounds with higher efficacy and less toxicity with hopes to improve outcomes for mUC.
Keywords: urothelial carcinoma, bladder cancer, second-line treatment, salvage, targeted therapy, gene expression profiling
1.Introduction
Bladder cancer (BC) is the most common type of urothelial carcinoma (UC) and the 9th most common cancer worldwide, with a high mortality rate [1]. The estimated worldwide incidence of BC was about 500,000 in 2018, with over 200,000 related deaths in that year [2]. In 2020 in the United States, BC is expected to be the 4th most common cancer among men [3]. Approximately 25% of UC patients have muscle-invasive BC (MIBC) or metastatic disease at the time of diagnosis, with poor 5-year overall survival (OS) rates of 35 and 5 percent, respectively [4]. Chemotherapy with cisplatin-based combination therapy (gemcitabine plus cisplatin [GC], methotrexate, vinblastine, doxorubicin,cisplatin [MVAC] and dose dense [DD]-MVAC) remains the initial standard treatment for cisplatin-eligible patients with mUC[5,6]. GC is preferred by many due to better safety profile. Over the course of the last several decades, the outcomes of metastatic (m) UC have remained poor, with patients receiving first-line cisplatin-based chemotherapy succumbing to their disease at a median of approximately 15 months [7-9]. The survival rates are worse in the presence of visceral metastasis or poor performance status (PS)[10-12]. Peri- operative cisplatin-based chemotherapy is also considered standard in patients with high-risk MIBC [13,14]. Approximately half of the mUC patients are not candidates for cisplatin-based combined chemotherapy [15]. UC affects older adults more commonly, with median age of 73 at the time of diagnosis [4], characterized by frequent comorbidities present among this age group leave many of these patients unfit to receive cisplatin and often unfit for any systemic chemotherapy.
The criteria published by a consensus working group defined being cisplatin-ineligible as having Eastern Cooperative Oncology Group (ECOG) PS 2 or greater, renal insufficiency (creatinine clearance less than 60 ml/min), heart failure (New York Heart Association class III or higher), severe peripheral neuropathy, or significant hearing loss [16].For long, combined chemotherapy using carboplatin in lieu of cisplatin has been used as the first-line treatment for patients who are unfit for cisplatin, with acceptable tolerability, yet lower objective response rate (ORR) of 30-40% and shorter median OS of 9-10 months [17-22]. Single-agent activity of various chemotherapy drugs has been explored in mUC trials, with no OS benefit [5,23-29]. Following the initial approval of immune checkpoint inhibitors (ICIs) as post-platinum-therapy, the approval of atezolizumaband pembrolizumab as new first-line treatment options for patients who are cisplatin-ineligible with high tumor PD-L1 expression or platinum-ineligible has offered new avenues [30- 32]. Despite these recent advances, the majority of UC patients progress through their first-line therapy for mUC or within a year of peri-operative chemotherapy for muscle-invasive UC, requiring salvage treatment.
The clinical outcomes are dismal in the salvage setting, and the outcomes of mUC continue to remain poor, with an unmet clinical need for more effective and tolerable therapies. Multiple study groups have attempted to define UC subtypes based on molecular characterization and common genetic predict response to treatment.Although BC and upper tract urothelial cancer (UTUC) share some genomic features, they exhibit certain differences. FGFR3 alterations, HRAS alterations and MSI high tumors have been observed more frequently in UTUC, while RB1 gene alterations are more common in BC [36]. Recent advances in understating the molecular biology of mUC have led to development of several novel agents. In this article, we will first review standard-of-care options in the salvage setting including historic chemotherapy regimens, as well as immune checkpoint inhibitor (ICI) immunotherapy for mUC (Table 1). In addition, we will discuss recently approved novel agents (Table 1), novel therapeutic targets Genetics behavioural and pathways under investigation in clinical studies and highlight some of the key ongoing clinical trials in each field.
2.Salvage Therapy
2.1 Chemotherapy
The choice of chemotherapy in the salvage setting is often limited by the upfront regimen used and the patient’s PS. The presence of liver metastasis, anemia and poor PS have been identified as pretreatment prognostic factors associated with worse outcomes and shorter OS in platinum-refractory mUC patients[37]. There are no standard second-line chemotherapy regimens after failure of first-line platinum-based therapy for mUC, although single agent therapy with vinflunine or taxanes have been historically used. A phase III randomized trial comparing the third generation vinca alkaloid vinflunine to best supportive care, demonstrated an ORR of 8.7 percent with vinflunine in mUC patients who had progressed on platinum-based chemotherapy. In the intention-to-treat population, survival benefit was not statistically significant (6.9 months with vinflunine vs 4.6 months with best supportive care, hazard ratio 0.88, 95% CI 0.69-1.12, p=0.287). However, in the trial-eligible population and after adjustment for prognostic factors, vinflunine showed a statistically significant OS benefit [38,39]. Vinflunine has been approved by the European Medicines Agency (EMA) and many other countries for the post-platinum treatment of mUC, but it was not approved in the United States. Multiple other chemotherapy agents including ifosfamide, pemetrexed, paclitaxel, nanoparticle- albumin-bound (nab)-paclitaxel, docetaxel, gemcitabine, and oxaliplatin have been studied in nonrandomized or retrospective studies in the second-line setting, showing poor ORR of 10-15% percent and short median OS of 7 to 9 months [23,27,40-46]. In a randomized phase II trial, nanoparticle albumin bound (nab)-paclitaxel every 3 weeks was not superior to paclitaxel every 3 weeks, although the ORR for paclitaxel was higher than previously reported (25%), suggesting that taxanes are a reasonable option in this setting[47]. Eribulin mesy late has demonstrated promising activity as a single agent or in combination with gemcitabine and may warrant further investigation [48]. A phase III trial of second-line treatment with gemcitabine and paclitaxel for mBC, conducted by the German Association of the Urological Oncology (AUO trial AB 20/99) reported an ORR of 40%. The median PFS and OS in the study were 3-4 months and 10 months, respectively [49]. In a pooled individual patient level analysis of 370 patients, those enrolled in trials of combination chemotherapy containing a taxane exhibited improved OS compared with patients enrolled in trials of single-agent taxane chemotherapy as salvage therapy for advanced UC (hazard ratio: 0.60; 95% confidence interval, 0.45-0.82; p=0.001) [50]. However, prospective randomized trial evaluation of salvage combination chemotherapy to validate these results has not been conducted.
2.2 Immune checkpoint inhibitors
After decades of discouraging clinical trial results, treatment choices for mUC in the salvage setting have broadened and advanced prominently. In the last few years, five ICIs, including anti-PD-1 inhibitors pembrolizumab and nivolumab, and anti-PD-L1 inhibitors atezolizumab, durvalumab, and avelumab, were approved by the FDA for the treatment of locally advanced or mUC progressing during or after platinum-based chemotherapy, regardless of PD-L1 expression [4,51-54]. In Europe, pembrolizumab, atezolizumab and nivolumabhave been approved in the same setting. The only ICI approved on the basis of a phase III trial is pembrolizumab. The KEYNOTE-045 phase III trial of pembrolizumab vs chemotherapy as second-line treatment for mUC, reported an OS benefit of approximately 3 months compared to chemotherapy [4,55]. The median OS was 10.1 months (95% CI 8.0–12.3 months) with pembrolizumaband 7.3 months (95% CI 6.1–8.1 months) with chemotherapy [hazard ratio (HR) 0.70; 95% CI 0.57–0.85; P < 0.001]. After a median follow-up of 27.7 months, the median 1- and 2-year OS rates were higher with pembrolizumab (44.2% and 26.9%, respectively) than chemotherapy (29.8% and 14.3%, respectively). However, PFS did not differ between treatment arms, although long-term, 1- and 2-year PFS rates were higher with pembrolizumab. Pembrolizumab also exhibited a higher ORR (21.1% versus 11.0%) and median duration of response (DOR) (not reached,range 1.6+ to 30.0+ months) vs. chemotherapy (4.4 months; range 1.4+ to 29.9+ months). Benefit was observed regardless of tumor PD-L1 expression. Pe mbrolizumab was associated with immune related adverse events in 19.5% of patients, most of which were mild. In contrast, pembrolizumab demonstrated lower rates of all toxicities (62.0% versus 90.6%) and severe toxicities (16.5% versus 50.2%) vs. chemotherapy.
The other ICIs are approved based on ORR of 15-20% in conjunction with durability of responses and similar toxicity rates [51,52,54,56]. Unfortunately, the IMvigor211 phase III trial did not demonstrate improved OS in post-platinum patients with high tumor immune cell (IC)-2/3 PD-L1 expression with atezolizumab vs. chemotherapy (median 11.1 months [95% CI 8.6-15.5; n=116] vs 10.6 months [8.4-12.2; n=118]; stratified hazard ratio [HR] 0.87, 95% CI 0.63-1.21; p=0.41), which precluded further formal hierarchical statistical analysis of the overall population [57]. Intriguingly, higher TMB was associated with improved OS in a retrospective analysis of the IMvigor211 trial [58]. A review of the data from different trials of the other approved ICIs in advanced UC also showed no clear association between the degree of PD-L1expression and ORR in post-platinum patients [59]. More recently, avelumab received FDA-approval based on the results of the JAVELIN-Bladder-100 phase III trial, which demonstrated improved OS (median 21.4 months vs. 14.3 months; hazard ratio for death, 0.69; 95% confidence interval [CI], 0.56 to 0.86; P = 0.001) as switch maintenance therapy in the regardless of tumor PD-L1 expression in patients with mUC who had stable or responding disease after platinum-based first line chemotherapy [60]. Better understanding of tumor biology is required for defining predictive markers in order to select the most appropriate treatment strategies upfront. Following their approval as post-platinum therapy, both pembrolizumaband atezolizumab were approved as first-line therapy for cisplatin-ineligible patients based on durable responses in 23% to 29% of patients in nonrandomized phase II trials.
Subsequently, data from ongoing first-line phase III trials of pembrolizumab (KEYNOTE-361) and atezolizumab (IMvigor130) for mUC showed compromised survival among patients with low PD-L1 expression in the ICI monotherapy arm compared to chemotherapy, restricting their use by the FDA and EMA as first-line options only for cisplatin-ineligible mUC patients who have high PD-L1 expression [61,62]. Preliminarily, the IMvigor130 phase III trial recently reported an improvement in progression-free survival (PFS) with the addition of atezolizumab to platinum-based first-line therapy in mUC (stratified HR 0.82 [0.70, 0.96], P = 0.007), although an extension of survival has not been identified yet [63,64]. The Phase III DANUBE first-line trial also did not meet the primary endpoints of improving OS compared to gemcitabine-platinum for durvalumabalone in patients with high tumor/immune cell PD-L1 or for durvalumabplus tremelimumab regardless of PD-L1 expression [65]. Data from other ongoing first-line phase III trials are awaited, which evaluate chemo-immunotherapy as well as single agent and combined ICI approaches. Furthermore, trials are evaluating various PD1/L1 inhibitors as neoadjuvant or adjuvant therapy with or without chemotherapy for MIBC based on promising phase II data and in combination with chemoradiation [66-69]. Despite a setback in the IMvigor010 phase III trial, which unfortunately did not report improved outcomes with adjuvant atezolizumab for MIBC following radical cystectomy, CHECKMATE274 was reported in a press release on September 24, 2020 to demonstrate improved disease-free survival for the overall as well as PD-L1 high expressing population as adjuvant therapy and we await the results of a trials evaluating pembrolizumab as adjuvant therapy. Moreover, pembrolizumab is already approved by the US FDA for BCG- unresponsive non-MIBC [70,71]. Thus, while we await results of all of the ongoing trials, PD1/L1 inhibitors are poised to be employed to treat multiple earlier stages of disease.
2.3 Erdafitinib
FGF/FGFRs are tyrosine kinase signaling pathways with key roles in the regulation of cell proliferation, migration, and survival [72-74]. FGFR alterations related to point mutations, gene amplification, or fusion can result in activation of the receptor and the downstream pathway, and have been identified in numerous malignancies, including breast, lung, gastric, hepatobiliary, melanoma, and bladder cancers[75-80]. Activating FGFR gene alterations are present in approximately 20% of mUC patients, and in about 37% of the urothelial cancers of the upper tract [3,81]. Hence, FGFR inhibitors have been vigorously investigated as therapeutic agents for rationally selected patients. Erdafitinib was the first targeted-therapy agent to receive FDA approval for the treatment of mUC [56]. Erdafitinibis an orally administered, potent and relatively specific small-molecule tyrosine kinase inhibitor targeting fibroblast growth factor receptor (FGFR) 1 through 4 [82]. In April 2019, erdafitinib was granted accelerated approval by the FDA for salvage therapy of locally advanced or mUC with FGFR2 or FGFR3 alterations after progression on at least one line of platinum-based chemotherapy [56]. A companion diagnostic assay was also approved that requires assessing the tumor for FGFR2/3 activating mutations/fusions using a simultaneously approved companion diagnostic test, therascreen FGFR RGQ RT-PCR Kit manufactured by Qiagen [56]. While the starting dose was 8 mg once daily day, a pharmacodynamically guided dose escalation to 9 mg was allowed if hyperphosphatemia was not attained. The results of the phase II trial of erdafitinib (BCL2001) showed an ORR of 40% (95% CI, 31-50), with 37% partial responses (PRs) and 3% complete responses (CRs). The median PFS and OS were 5.5 months and 13.8 months, respectively.
FGFR3 mutation was associated with a higher ORR (49%), while the ORR was lower in those with FGFR2/3 fusion (16%) [82]. Of the study patients (n=99), those previously treated with ICI (n=22) had an ORR of 59%, of whom only one patient had exhibited previous response to ICI (21 of 22). Notably, the response rate was not significantly affected by the number of previous treatment lines or presence of visceral metastasis. Despite the remarkable ORR of erdafitinib, the DOR was modest, with a median of 5.6 months. Toxicity with erdafitinibis notable, and can be a limiting factor. According to the phase II trial, all of the participants had some degree of toxicity, with almost half experiencing grade 3/4 AEs, including hyponatremia, stomatitis, asthenia, palmar-plantar erythrodysesthesia syndrome, as well as eye toxicities. Drug-related toxicities resulted in dose reduction or interruption frequently (53% and 68%, respectively). Treatment was discontinued permanently in 13% of the patients due to AEs, mostly related to ocular toxicity. Grade 3/4 eye AEs, including retinal pigment epithelial detachment or central serous retinopathy occurred in 10% of the patients, plus more frequent toxicities of lower grades. Hence, periodic ophthalmological exams for the patients, on a monthly basis during the first 4 months, and every 3 months thereafter have been recommended [82].Increase in phosphate levels is a common issue with erdafitinib (77%), resulting from inhibition of the FGFR/FGF23-Klotho axis in the kidney tubules and disruption of phosphate homeostasis [83].However,in conjunction with a low phosphate diet, and pharmacologic intervention by deploying phosphate binders, severe hyperphosphatemia occurred only in 2% of the patients in the phase II trial [82].
2.4 Enfortumab vedotin
EV is an ADC consisting of a monoclonal antibody targeting nectin-4 bound to monomethyl auristatin E (MMAE), a microtubule-disrupting agent leading to mitosis arrest and cell death. Nectin-4 is a cell adhesion molecule with high expression on the surface of urothelial cells and other solid tumors, including breast, lung, pancreatic, and esophageal cancer [84]. EV was granted breakthrough therapy designation by the FDA for mUC after progression on ICI, based on the interim results from a phase I study of EV for mUC after checkpoint immunotherapy [85]. The phase II trial results of EV, EV-201, demonstrated an ORR of 44% (28% PRs and 12% CRs) with EV monotherapy following prior platinum-based chemotherapy and ICI therapy [86,87]. Robust responses were seen inpatients with no previous response to ICI (41%) and those with liver metastases (38%).The toxicity profile of EV was acceptable with the most common grade 3 AEs reported as fatigue, anemia, and neutropenia. The most common treatment-related adverse events in the pivotal phase II trial were fatigue (50%), peripheral neuropathy (50%), alopecia (49%), rash (48%), decreased appetite (44%), and dysgeusia (40%)[87].Treatment was discontinued in 12% of the patients due to AEs [86]. The results of the long-term follow-up of EV-201 study, presented at ESMO in September 2020, showed a median OS of 12.4 months (95% CI, 9.46- 15.57), with OS rates of 50% and 34% at 12 and 18 months, respectively [88].
The median follow-up duration was 22.3 months. The confirmatory phase III trial (EV-301) of second-line therapy with EV vs salvage chemotherapy, including vinflunine or taxane, was recently reported to show improved OS(Hazard Ratio [HR]=0.70; [95% CI: 0.56, 0.89]; p=0.001) and PFS ((HR=0.61 [95% CI: 0.50, 0.75]; p<0.00001) in a press release on September 18, 2020 (NCT03474107)[89]. Moreover, results from the second cohort of patients in the phase single-arm EV-201 clinical trial were announced in a press release on October 12, 2020. The cohort evaluated EV for cisplatin-ineligible patients had been previously treated with a PD-1/L1 inhibitor but had not received a platinum-containing chemotherapy. Results showed a 52 percent ORR by blinded independent central review and a median duration of response of 10.9 months. While carboplatin-based chemotherapy may be considered the current community standard for this group of patients, EV represents an emerging option in such patients based on these results. Additionally, EV is also being tested as a phase Ib trial in combinations with PD-1 inhibitor pembrolizumab and chemotherapy (NCT03288545)[90]. Preliminarily the combination of EV and pembrolizumab demonstrated feasibility and promising activity as first line therapy for cisplatin- ineligible mUC [90,91]. Ongoing evaluation of the combination of first line EV with chemotherapy and/or pembrolizumabis ongoing in several cohorts of the EV103 trial.
2.5 Investigational agents as salvage therapy
2.5.1Immune checkpoint inhibitors
Ipilimumab may have modest activity as a single agent and in combination with first line GC, although the combination with nivolumab has appeared the most promising avenue for use in mUC [92]. A phase I/II study, CheckMate 032 (NCT01928394), evaluated nivolumab monotherapy and 2 dose levels of combined nivolumab andipilimumabin pretreated advanced/metastatic tumors, including platinum- resistant mUC [93,94]. The arm receiving ipilimumab (3 mg/kg) plus nivolumab (1 mg/kg),had the highest ORR (38% overall, and 58% among PD-L1-high tumors) and median PFS of 4.9 months vs. the nivolumab (3 mg/kg) plus ipilimumab (1 mg/kg) and nivolumabalone (median PFS 2.6 and 2.8 months,respectively). These results led to conduction of a phase III trial comparing the ipilimumab (3mg/kg) plus nivolumab (1mg/kg) combination vs chemotherapy for upfront treatment of mUC (NCT03036098). Another trial (NCT02261220), phase I study of PD-L-1 inhibitor durvalumab combined with CTLA-4 inhibitor for chemotherapy-pretreated patients, showed an ORR of 20.8%, with an ORR of 29.4% and median OS of 18.9 months in PD-L1-high patients [95]. This study was followed by a phase III trial, DANUBE (NCT02516241), which unfortunately could not demonstrate improved outcomes as mentioned earlier. NKTR-214, a CD122-biased agonist designed to provide sustained signaling through the IL-2 βγ receptor demonstrated promising activity in combination with nivolumabin the first-line setting and is being developed in the neoadjuvant and first line settings for cisplatin-ineligible patients in phase III trials [96]. A phase Ib/II umbrella study (MORPHEUS mUC) is assessing the efficacy and safety of different combinations of PD-L1 inhibitor atezolizumab with various immunotherapy-based agents, including enfortumab-vedotin (EV), PARP inhibitor niraparib, anti-CD47 antibody Hu5F9-G4, anti-TIGIT tiragolumab, DPP-4 inhibitor linagliptin, and anti-IL-6 tocilizumab (NCT03869190).
2.5.2FGFR inhibitors
Various other FGFR inhibitors have been tested via phase I and II clinical trials with ORRs of around 25%- 35% including rogaratinib, pemigatinib, Debio-1347 and infigratinib [3,97-102]. In addition, there are multiple ongoing trials studying FGFR pathway inhibition as single agent or in combination with other drugs. Vofatamab (B-701), an anti-FGFR3 monoclonal antibody,is being tested both as monotherapy,and combined with pembrolizumab for mUC. Vofatamab has the potential to block the activation of the FGF/FGFR3 pathway by binding to both the wildtype (WT) FGFR3 as well as the activated FGFR3 in those with gene mutation or fusion (mut/fus). FIERCE-022 is a phase Ib/II trial studying vofatamabalone for 2 weeks (at 25 mg/kg) followed by vofatamab combined with pembrolizumab (200 mg every 3 weeks) for mUC patients with failure of prior chemotherapy. Phase III trials are attempting to confirm a role for erdafitinib and rogaratinib in molecular lyselected post-platinum metastatic mUC patients (Table 2). The results of the first interim analysis have reported promising activity in the WT as well as altered group[86,103].
2.5.3 Sacituzumab govitecan
Sacituzumab govitecan (SG) is an ADC consisting of anti-Trop-2 monoclonal antibody connected to SN38, the active metabolite of irinotecan [104]. Trop-2 is a transmembrane glycoprotein with high expression in UC as well as other solid tumors including prostate, triple-negative breast cancer, colonic, gastric, pancreatic, and lung cancer [105]. In the mUC cohort of a phase 1/2 trial, 45 evaluable post-platinum mUC patients received SG [106]. The ORR was 31% and the median DOR was 12.9 months. Grade ≥3 adverse events in ≥5% of patients were neutropenia (38%), anemia (13%), hypophosphatemia (11%), diarrhea (9%), fatigue (9%), and febrile neutropenia (7%). The median PFS was 7.3 months and median OS was 16.3 months. Thereafter the TROPHY-U-01 phase II trial was conducted enrolling patients with progression after both platinum-based chemotherapy and PD1/L1 inhibitors. Preliminarily, an interim analysis of 35 patients in cohort 1 of the TROPHY-U-01 trial reported an ORR of 29%, including responses in 25% with liver metastasis [107].The final results of this cohort (n=113) were presented at ESMO virtual congress in September 2020, which reported an ORR of 27% with a CR and PR in 5% and 22%, respectively. The median DOR was 5.9 months, and the median PFS and OS were 5.4 months and 10.5 months, respectively [108]. The phase III confirmatory trial, TROPiCS-04, is currently ongoing (NCT04527991).
2.5.4 HER-2 inhibitors
Human epidermal growth-factor receptor- 2(HER-2) alterations have been identified as markers with predictive and prognostic values in a variety of tumors, making them interesting targets for treatment[109-111]. In UC, HER-2 amplifications have been reported with various frequencies by different studies, and they have been reported in higher frequencies in the micropapillary subtype[112]. The Cancer Genome Atlas (TCGA) data reported HER-2 amplification or mutations in approximately 10% of UCs [113-116]. The combination of trastuzumab, paclitaxel, carboplatin and gemcitabine yielded significant activity in a phase II trial as first-line therapy in mUC with HER2 overexpression by immunohistochemistry (IHC), gene amplification and/or elevated serum HER2.[117] In this trial, 31 (70%) of 44 patients responded and the median time to progression and survival were 9.3 and 14.1 months, respectively. Unfortunately, a randomized trial did not demonstrate improved outcomes for trastuzumab combined with platinum- based first-line chemotherapy in patients bearing HER2 overexpressing (3+ by IHC or 2+ by IHC and FISH+) mUC.[118] A combination of pertuzumaband trastuzumab without chemotherapy exhibited modest activity in previously treated HER2-amplified mUC.[119] Moreover, improved outcomes were not observed in a phase III trial that examined second-line switch maintenance lapatinib inpatients with HER1 or HER2 protein overexpressing tumors [120,121]. Afatinib, a potent irreversible pan-HER inhibitor displayed promising activity in those with HER genomic alterations. Five of 6 patients with HER2 and/or HER3 alterations achieved prolonged PFS compared to none of 15 patients without alterations.
A larger basket study is evaluating afatinib for genomically selected patients (Table 2) [122]. In another basket trial, the HER kinase inhibitor neratinib failed to achieve a partial response in any of 16 patients with ERBB2 mutant UC [123]. While trastuzumab (T)-DM1 has promising activity in preclinical models of HER2 over expressing bladder cancer, clinical data are unavailable [124]. Recently, novel HER2 targeting ADCs and small molecule inhibitors have demonstrated promising activity. RC48-ADC, an anti-HER-2 ADC incorporating MMAE as the toxic payload reported a confirmed ORR of 51% and median PFS of 6.9 months in 43 patients with HER-2 overexpressing (IHC 2+ or 3+) mUC. [125]. Activity was preserved in those with visceral metastasis and following prior ICI therapy. The major grade ≥3 toxicities were hypoesthesia and neutropenia. [125]. Trastuzumab deruxtecan, DS- 8201a, is another ADC formed of anti HER-2 antibody trastuzumab bound to topoisomerase-I inhibitor deruxtecan. This compound showed encouraging results in HER-2 positive breast and gastric cancer patients previously [126,127]. A phase Ib trial (NCT03523572) is studying DS-8201a in combination with nivolumab for HER-2 IHC+ mUC and breast cancer. PRS-343 is a novel bispecific molecule being studied in HER-2 positive mUC patients. PRS-343 was formed by recombinant fusion of an anti-HER-2 antibody with a CD137 (4-1BB)-specific molecule, a costimulatory receptor expressed on activated T cells and natural killer (NK) cells [128]. PRS-343 is currently being studied in phase I trials (NCT03330561 and NCT03650348), as monotherapy and in combination with atezolizumab.
2.5.5 VEGF/VEGFR inhibitors
Vascular endothelial growth factor (VEGF) pathway inhibition has been tested across different stages of UC. VEGF receptor (VEGFR) TKIs such as sunitinib, cabozantinib, pazopanib, regorafenibandsorafenib have shown modest activity as monotherapy [ 129-131]. The combination of VEGF TKIs with chemotherapy has unfortunately been either prohibitively toxic (e.g. pazopanib plus paclitaxel, GC plus sunitinib) or not led to an increment (docetaxel plus vandetanib). The combination of VEGF and PD1/L1 inhibitors may be synergistic owing to the immune modulating effects of VEGF inhibition. Cabozantinib demonstrated promising immune modulating activity and was combined with nivolumabalone or with ipilimumabin pre-treated patients [132]. The combination of lenvatiniband pembrolizumab as well as cabozantinib and atezolizumabhave also appeared active inpatients with advanced UC [133,134].Bevacizumab, a VEGF-targeted monoclonal antibody, has been tested in combination with chemotherapy via phase II and III trials. In a phase II trial combining bevacizumab with GC, CR was seen in 19% and PR in 53% (ORR=72%) in conjunction with median PFS and OS of 8.2 and 19.1 months,respectively. Despite the impressive outcomes in the phase II trial in combination with GC, the randomized double-blind phase III trial, CALGB 90601, did not show an OS benefit (HR 0.87, p=0.17), despite the modest improvement of PFS (HR 0.77, p=0.0074) [135,136]. A phase III trial evaluating docetaxel in combination with ramucirumab, an IgG1 anti-VEGFR-2 antibody, versus placebo for platinum-refractory mUC, demonstrated a modest PFS benefit (median 5.4 vs 2.8 months, HR 0.39; p=0·0002); however, there was no OS benefit [137,138].
2.5.6 PI3K/AKT/mTOR inhibitors
The phosphatidylinositol 3-kinase (PI3K) pathway is involved in cell proliferation, migration and survival [139]. Alterations resulting in activation of the PI3K/AKT/mTOR pathway are common in mUC, making them potential therapeutic targets [140,141]. PK3CA, TSC1 and TSC2, which constitute the mTOR regulatory tuberous sclerosis complex are mutated in 15-20% of advanced bladder cancer [142,143]. Everolimus has anecdotally exhibited extremely durable responses in a proportion of patients with somatic genomic alterations [144] The results of a phase I study using combination of mTOR inhibitor everolimus and VEGR-TKI pazopanib in previously treated mUC patients showed an extreme response in one patient [145]. As a result, an expansion cohort of mUC patients was added to the trial which showed an ORR of 21%. Of the 12 mUC patients who had responded, 5 had mTOR pathway alterations [146]. A phase II trial of paclitaxel and everolimus combination as salvage therapy in mUC (AUO Trial AB 35/09)did not show significant efficacy [147]. Vistusertib, a small-molecule mTORC1 and mTORC2 inhibitor, was administered in combination with durvalumab under the phase Ib umbrella BISCAY study enrolling previously treated mUC patients harboring somatic genomic alterations of RICTOR or TSC1/2 (NCT02546661) [148]. Unfortunately, the ORR of 24.1% did not meet the predefined ORR of interest, which was 50%. Recently, the result of the phase 2 study of PI3K inhibitor buparlisib demonstrated poor activity in somatic-TSC 1 altered patients [149]. Poly ADP ribose polymerase (PARP) inhibitors have shown activity in tumors with DNA damage repair (DDR) gene alterations. A number of ongoing trials are investigating the activity of PARP inhibitors alone, or in combination with ICI for mUC with and without DDR alterations. The rationale for this combination is the hypothesized increase in TMB and neoantigens with PARP inhibition. [150-152]. Disappointing results were presented for rucaparib in the ATLAS trial, which did not yield objective responses by RECIST1.1 although some minor regressions were observed in patients with previously treated mUC with or without somatic genomic homologous recombination deficiency (HRD) status [153]. The aforementioned BISCAY trial included a cohort receiving olaparib in combination with durvalumab following prior chemotherapy, which did not attain the predefined 50% ORR of interest [148]. A phase II open-label randomized trial is investigating the effect of niraparib plus best supportive care versus best supportive care as maintenance therapy after a first-line platinum-based chemotherapy (NCT03945084).
3.Conclusion
The landscape of salvage therapy has transformed and advanced significantly in the recent years with the addition of new therapeutic options following prior platinum-based chemotherapy. The approval of 5 PD-1/PD-L1 ICIs and EV as well as erdafitinib for selected patients with somatic FGFR3/2 genomic alterations has supplanted historically used chemotherapy and transformed the therapeutic landscape.There are a variety of other agents directed against different molecules and pathways currently being tested (Figure 1). Despite these advances, mUC remains an incurable disease. Hence, a better understanding of tumor biology is necessary to tackle this molecularly heterogeneous malignancy.
4.Expert Opinion
Given the emergence of multiple agents to treat mUC, the optimal sequencing of agents needs study, especially inpatients harboring FGFR3/2 altered tumors. In particular, the sequencing of post-platinum PD1/L1 inhibitors and erdafitinib as well the sequencing of EV anderdafitinib requires study. [65]. FGFR alterations are more frequent in luminal-like UCs, which are associated with lower T cell infiltration and lower response rates to ICIs [51,154,155]. These findings had previously proposed FGFR alteration as a predictive marker of resistance to ICIs [82]. Indeed only 1 of 22 patients responded to prior ICI before they went on to receive erdafitinib and generally responded with an ORR of 59%. However, a recent study of the correlative data from two trials for mUC (IMVigor 210 using atezolizumaband CheckMate 275 using nivolumab) showed similar ORRs of ~25% to ICI among FGFR3mut and FGFR3WT patients [154,156,157]. Moreover, the sequencing of carboplatin-based chemotherapy and PD1/L1 inhibitors in patients with PD-L1 expressing mUC requires study. Notably, the DANUBE trial did not demonstrate improved OS when using first line durvalumab compared to gemcitabine-platinum for mUC patients harboring PD-L1 high-expressing tumors. Precision medicine strategies have made inroads with the approval of first line pembrolizumaband atezolizumab for cisplatin-ineligible mUC with high tumor PD-L1 expression and erdafitinib for post- platinum patients with somatic FGFR3/2 genomic alterations. However, biomarkers are not in the clinic to select patients for platinum-based chemotherapy, EV or historical chemotherapeutic agents, vinflunine and taxanes. While data suggest the association of DNA damage repair (DDR) alterations with vulnerability to cisplatin-based chemotherapy and the association of TMB with vulnerability to ICIs, these biomarkers are not validated for primetime use in the clinic [58,158].
Preliminary data also suggest that a combination of genomic (TMB) and clinical factors (sites of metastasis, peripheral blood neutrophil-lymphocyte ratio) may enhance the ability to predict benefit from ICIs as opposed to other agents [159]. Continued efforts to identify biomarkers of efficacy and toxicities need to continue to enhance the therapeutic index as well as cost-efficacy of these regimens. Advances in precision oncology and discovery of new therapeutics requires thorough molecular analyses of the tumor and host in preclinical and clinical studies. It is also important that trials that have reported disappointing or underwhelming results inform future drug development. There have been some notable failures in clinical trials despite the supportive evidence for efficacy in preclinical studies. In UC, Epidermal growth factor receptor (EGFR) overexpression is associated with higher grade and muscle invasion of the tumor as well as recurrence rate and OS [115,160]. Unfortunately, trials of EGFR-inhibitors, including cetuximab, panitumumab, gefitinib, and lapatinib for mUC, have shown poor or no activity [161-163]. The cyclin dependent kinase 4 and 6 (CDK4/6) – retinoblastoma (Rb) pathway controls cell cycle transition from G1 to S phase and CDK4/6 inhibition has been shown to induce tumor cell cycle arrest through restoration of Rb function [164,165]. Unfortunately, a phase II trial of CDK4/6 inhibitor palbociclib in the salvage setting did not show significant activity inpatients selected by IHC [166].
While alterations in chromatin-modifying genes, including CREBBP and EP300, which can lead to dysregulated acetylation, are seen in approximately a third of UC patients, disappointing results were observed with mocetinostat in selected patients and vorinostat in unselected patients [167,168]. The combination of ICIs and epigenetic modulation to induce viral mimicry and neoantigen expression and reprogramming of the previously exhausted T cells was hypothesized to augment antitumor activity. Based on this hypothesis, a phase I/II trial (NCT03179943) of atezolizumab combined with guadecitabine, a DNA methyl transferase inhibitor following prior ICI is ongoing. These results suggest that a more thorough understanding of tumor biology in preclinical models that more optimally reflect human subjects is critically important. Multiple efforts to improve outcomes with salvage therapy for mUC have already borne fruit with the arrival and regulatory approvals of PD1/L1 inhibitors, EV and erdafitinib as single agent salvage therapy. examining Mature data from ongoing first line trials evaluating the combination of platinum-based chemotherapy and PD1/L1 inhibitors, the combination ICI approach and switch maintenance avelumab are awaited, which will impact salvage therapy development. Phase III first line trials are reading out with modest PFS benefit reported for the combination of platinum-based chemotherapy and atezolizumab regardless of PD-L1 status. However, OS benefit needs to be demonstrated for bringing the chemo-immunotherapy strategy to the clinic. Thus, the currently employed paradigm for first line therapy has not yet changed.
Moreover, results of other ongoing trials evaluating the combination of other PD1/L1 inhibitors (pembrolizumab,nivolumab, durvalumab) with platinum-based chemotherapy are awaited (Table 2). Although the DANUBE trial disappointed with failure of combination durvalumab- tremelimumaband durvalumab alone to yield OS benefits in the unselected and PD-L1 high populations, respectively, we are awaiting data from a phase III trial investigating ipilimumab-nivolumabin the same setting. Indeed, if these trials usher in a major role for combination chemoimmunotherapy followed by continued PD1/L1 inhibitor or switch-maintenance PD1/L1 inhibitors, the combination of novel tolerable agents with the post-platinum PD1/L1 inhibitor may be an attractive space to develop combinations while maintaining an optimal therapeutic index. The next phase of new salvage therapy drug development will consist of a multipronged effort to 1) combine novel agents with proven and different mechanisms of activity in order to overcome resistance, including the combination of approved (EV) and emerging ADCs (SG) with ICIs, the combination of FGFR and VEGF inhibitors with ICIs, the combination of PD1/L1 inhibitors with other checkpoint inhibitors and the combination of targeted agents in rationally selected patients 2) continuing advances invalidation of predictive markers to optimally sequence novel agents guided by non-invasive biomarkers before each new line of therapy and 3) assess mechanisms of resistance and tumor biology with metastatic tumor biopsy protocols and rapid autopsies to inform rational drug development [150,169].
There may be a rationale to enhance the activity of FGFR inhibitors by combining with other agents guided by tumor biology. Using gene expression data of the mut and WT cells, the study by Wang et al. demonstrated similar tumor mutation burden (TMB) between the mut and WT groups. However, the expression of TGF-B-related genes was significantly lower in the FGFR3mut cells compared with FGFR3WT cells. This generates a hypothesis concerning potential beneficial immunomodulatory effects of ICIs in combination with FGFR3 inhibition in those with FGFR3 alterations [156]. Indeed, the combination of FGFR inhibition and PD1/PD-L1 inhibition is being evaluated, given the potential for additive or synergistic efficacy. Rogaratinibis also being evaluated in combination with copanlisib, a PIK3CA inhibitor, in a phase I trial enrolling metastatic UC given evidence for the role of PIK3CA in promoting resistance to FGFR inhibition (Table). Similarly, the combination of EV and pembrolizumab appeared promising as first line therapy for cisplatin-ineligible patients with mUC Polymer bioregeneration [170]. The ORR among 45 patients was 73.3% including 15.6% complete responses (CRs). The median DOR was not reached (range 1.2 to 12.9+ months) and the median PFS was 12.3 months. The US FDA has granted a breakthrough therapy designation for this combination. Meanwhile phase III evaluation of this combination is ongoing in the 3-arm EV-302 trial, which compares the combination of EV and pembrolizumab with or without platinum vs. platinum-gemcitabine combination chemotherapy as first line therapy for mUC. A number of studies are using novel agents in combination with ICI predicated on augmenting the immunogenicity of the tumor or tackling the immunosuppressed tumor microenvironment to transform cold or immune-excluded tumors to hot tumors.
In addition to NKTR-214, other ICIs such as TIM-3 and LAG-3 antagonists, and immune agonists such as OX40 agonists are also undergoing early clinical development. Agents that modulate metabolites and favorably alter the immune tumor microenvironment such as arginase and adenosine inhibitors are also undergoing early development. Other novel agents include a variety of modalities such as vaccines (off the shelf or customized), plasmids expressing immunogenic DNA (e.g. INO-5401, comprising 3 synthetic plasmids against Wilms’ tumor gene-1, prostate-specific membrane antigen and human telomerase reverse transcriptase antigen, and INO-9012, a synthetic plasmid expressing human interleukin (IL)-12 and intratumoral mRNA injection (e.g. mRNA-2752, a lipid nanoparticle encapsulating mRNAs encoding for the co-stimulatory tumor necrosis factor ligand superfamily member 4, OX40 Ligand, interleukin [IL]-23 and IL-36γ). VEGF inhibitors may enhance the immune system in a number of ways including tumor ingress of immune cells and clearance of immunosuppressive MDSCs and Tregs. Multiple combinations of VEGF inhibitors and ICIs are being evaluated including cabozantinib and axitinib (Table). Tyrosine kinase inhibitors targeting the TAM (Tyro, Axl, MER) kinases appear to induce an immune-promoting state, which led to a trial evaluating the combination of sitravatinib with nivolumab. Preliminarily, promising activity was observed with confirmed responses in 6 of 22 mUC patients following progression on a prior ICI (NCT03606174).
Past unsuccessful efforts to enhance the immune microenvironment need to inform future drug development, e.g. the combination of acalabrutinib, a BTK inhibitor shown preclinically to suppress MDSCs, with pembrolizumab did not unfortunately improve outcomes in unselected post-platinum patients [171]. The aforementioned ATLAS trial could not demonstrate significant activity for one of the PARP inhibitors, rucaparib, even in post-platinum mUC patients selected for HRD. The effort to combine a PD-L1 inhibitor (durvalumab) with an FGFR inhibitor (AZD4547) in selected post-platinum mUC patients with genomic alterations in the BISCAY trial did not attain the predefined 50% ORR of interest [148]. In contrast, the combination of neoadjuvant durvalumab andolaparib appeared promising with 10 of 20 patients attaining pCR [172]. Given Akt inhibitor the molecular heterogeneity of the disease, targeting molecular subsets of the disease in an earlier line of therapy guided by preclinical data may be warranted. While historically used HER2 inhibitors have not been successful, more potent novel inhibitors may warrant investigation. In this context, afatinib, trastuzumab deruxtecan in combination with nivolumab and PRS-343 are undergoing investigation. Tucatinib, a potent and highly selective HER2 tyrosine kinase inhibitor (TKI), and margetuximab, an ADC optimized for antibody dependent cell-mediated cytotoxicity (ADCC) may also warrant study based on promising activity in heavily pretreated HER2 positive patients with breast cancer.
Moreover, novel therapeutic targets are being discovered, which may be relevant in rationally selected patients, e.g. PAK4 kinases appeared to drive a subset of cancers and may constitute a therapeutically actionable target [173]. Finally, cytotoxic agents may continue to play a role given that mUC is generally a highly proliferative and aggressive malignancy. Eribulin will undergo phase III evaluation as single agent or in combination with gemcitabine as salvage therapy in a U.S. co-operative group trial. Interestingly, the combination of post-platinum pembrolizumab combined with weekly nab-paclitaxel showed promising activity in a phase II trial[174]. Thus, the rational development of selected cytotoxic agents with favorable therapeutic indices may be warranted. Thus, a multipronged effort to understand tumor biology, which in turn guides rational drug development will allow us to make further advances in the salvage therapy of mUC and eventually optimize first line therapy.
