*3.1. Enfortumab Vedotin (Padcev; Astellas; Tokyo, Japan; and Seattle Genetics; Bothell, WA, USA)*

EV is an ADC that targets Nectin-4 that is overexpressed on the surface of bladder tumor cells. EV is conjugated to the microtubule inhibitor monomethyl auristatin E that causes G2/M cell cycle arrest and results in apoptosis. [55] Nectins are involved in cellular adhesion, migrations and polarization [56]. IHC analysis showed Nectin-4 to be overexpressed in 93% of metastatic urothelial tumor specimens [55]. In contrast, 294 normal tissue specimens representing 36 human organs showed homogeneous weak to moderate staining. Nectin-4 expression via IHC staining was determined by the H-score. The H-score was calculated by summing the products of the staining intensity (score of 0–3) multiplied by the percentage of cells (0–100) stained in a given field of tumor tissue [55]. Specimens were then classified as negative (H-score 0–14), weak (H-score 15–99), moderate (H-score 100–199), and strong (H-score 200–300). Thus, Nectin-4 is an attractive target due to its preferential overexpression in bladder cancer relative to normal tissues.

### 3.1.1. EV-101 Trial

EV was approved in the United States in December 2019 based on the results from the phase I and II EV-101 and EV-201 clinical trials, respectively (extensively reported in [57]) (Table 1) [58]. Of note, patients in the EV-101 study who had previously received ICI therapy had an ORR of 42% (95% CI, 31.2–52.5%). In addition, patients with high tumor burdens such as liver metastases had a 36% (95% CI, 20.4–54.9%) ORR [59].

#### 3.1.2. EV-201 Trial

In the phase II EV-201 trial, 125 patients with locally advanced or metastatic bladder cancer who were previously treated with platinum-containing chemotherapy or ICI therapy were treated with EV. Tumor expression levels of Nectin-4 and PD-L1 were evaluated. Nectin-4 expression levels were evaluated and scored as previously described [55]. The median Nectin-4 expression level was H-score = 290 (14–300) and, hence, all patient tumors evaluated were positive for Nectin-4 and, they were considered as having 'strong' expression. PD-L1 expression was scored as previously performed for pembrolizumab with tumors being classified as positive with a score ≥10 [42]. The proportion of patients with <10 and ≥10 PD-L1 scores was 65% and 35%, respectively.

At the median follow-up time point of 10.2 (0.5–16.5) months the ORR was 44% (95% CI, 35.1–53.2%), including 12% with complete responses regardless of PD-L1 status (Table 1) [34]. The median DOR was 7.6 (4.9–7.5) months. The median OS was 11.7 (9.1—not reached) months. Hence, these patient outcomes are reflective for tumors that had 'strong' Nectin-4 expression. In contrast, in the PD-L1 subgroups the ORRs were 47% (95% CI, 36–59.1%) and 36% (95% CI, 21.6–52%) for TPS < 10 and TPS ≥ 10 PD-L1 expression scores, respectively. This indicated that patients responded regardless of PD-L1 expression. This study demonstrated, (i) the importance of Nectin-4 as a biological target for bladder cancer, relative to PD-L1, and (ii) EV has the potential to significantly extend the lives of patients, including those who failed ICI treatment.

EV is currently being evaluated in the phase III EV-301 clinical trial (NCT02091999). In this global study, approximately 550 patients are being randomized to receive EV or investigator's choice of docetaxel, paclitaxel, or vinflunine [60]. There is also currently active recruitment for a phase II study (NCT03288545) to evaluate EV alone and in combination with various anticancer therapies, including the ICI pembrolizumab [61,62]. Preliminary results have shown an ORR of 71% with the combination of EV plus pembrolizumab in 45 cisplatin-ineligible patients. The available data for DOR, PFS, and OS are not yet mature. Taken together, these clinical studies reveal EV has the potential to significantly extend the lives of patients who fail ICI treatment and has the potential to synergize patient response with ICI therapy.

#### *3.2. Sacituzumab Govitecan (Trodelvy; Immunomedics; Morris Plains, NJ, USA)*

SG is an ADC that targets Trop-2 that is overexpressed on the surface of bladder tumor cells. SG is conjugated to the topoisomerase I inhibitor SN-38 [63]. It is currently approved for use in patients with triple-negative breast cancer who have received at least two prior forms of chemotherapy. Trop-2 is a transmembrane glycoprotein, which participates in cellular self-renewal, invasion, proliferation and survival and overexpressed in multiple solid tumors, including bladder cancer [64]. An IHC analysis showed Trop-2 was generally overexpressed in bladder tumor tissue with little expression detected in the corresponding normal tissue [65]. The method in which Trop-2 expression level was determined was not provided.

#### IMMU-132 Trial

SG was evaluated in the phase I/II IMMU-132 clinical trial in patients with advanced bladder cancer that received prior platinum-based treatment (Table 1) [35]. Patient tumors were determined as positive if >10% of tumor cells had anti-Trop-2 staining. Expression was scored as 3+ (strong), 2+ (moderate), and 1+ (weak) [35]. Tumors with <10% tumor cells that stained for Trop-2 were considered Trop-2-negative tumors. SG treatment resulted in an ORR of 31%, with two complete and 12 partial responses out of 45 patients. In a patient cohort previously treated with ICIs, the ORR was 23% (4/17). A single-arm, open-label, global TROPHY U-01 phase II trial is currently ongoing to evaluate SG in advanced bladder cancer (NCT03547973). Interim results from 35 patients from the 100-patient cohort of cisplatin-eligible patients who have also previously received ICI therapy and platinum-containing chemotherapy showed an ORR of 28% [66]. Based on these results, the FDA granted Immunomedics request for fast track designation in order to make SG available as rapidly as possible. It is not known whether there was a correlation in patient responses with respect to Trop-2 expression levels.

#### **4. Erdafitinib (Balversa; Janssen Pharmaceuticals; Beerse, Belgium)**

Bladder cancer is the third most common mutated malignancy and has the strongest association to fibroblast growth factor receptors (FGFRs) 1–4 gene mutations relative to all other cancer types [67,68]. Moreover, FGFR mutational aberrations occur in >50% of all bladder cancer cases [68]. Interestingly, FGFR3 mutations occur in 60% of invasive bladder tumors and it is a poor prognostic marker [69].

FGFRs represent a family of tyrosine kinases found on the surface of normal cells. There are currently four recognized receptor isoforms, which bind corresponding ligands, and leads to receptor dimerization and phosphorylation [68]. Ligand binding and dimerization results in downstream signaling and expression of several gene products that function to promote cell survival and proliferation. Gene mutations including gene amplification combine to promote cell growth beyond normal limits and results in the development of cancer. Further abnormal FGFR mechanisms that promote bladder cancer is nicely reviewed in Roubal et al. [70].

#### *BLC2001 Trial*

Erdafitinib is a pan-FGFR inhibitor and exerts is action by binding to and blocking FGFR phosphorylation and signaling, which decreases cell viability, particularly in tumor cells with FGFR genetic alterations. It was approved on 12 April 2019 for use in patients with locally advanced or metastatic bladder cancer, with susceptible FGFR3 or FGFR2 genetic alterations, that has progressed during or following platinum-containing chemotherapy, including within 12 months of neoadjuvant or adjuvant platinum-containing chemotherapy [71]. Approval was based on its performance in the phase II BLC2001 clinical trial (Table 1) [36]. Patients with locally advanced and unresectable or metastatic bladder cancer, with prespecified FGFR alterations, and who had been previously treated with platinum-containing chemotherapy were enrolled into the study. Importantly, a proportion of enrolled patients also received prior ICI therapy. Erdafitinib treatment resulted in an impressive ORR of 40% (95% CI, 31–50%). The median OS was 13.8 (9.8—not reached) months. In addition, at 1 year, 19% (11–29%) of patients continued to respond to treatment. Interestingly, Patients with FGFR3 mutations had the highest ORR at 49%. In contrast, patients with FGFR fusions had the lowest ORR at 16%. This indicates that erdafitinib can serve patients more effectively with tumors that contain FGFR mutations as opposed to fusions. In addition, erdafitinib can improve the outcomes of patients who previously received ICI therapy.

These targeted drugs, finally, are clinical breakthroughs that demonstrate molecular targets can result in precision therapeutics that are highly effective against metastatic bladder cancer. In addition, these drugs can also improve outcomes of patients that don't respond or relapse after receiving ICI therapy.

#### **5. Adverse Events**

Comparative percentages between the described ICIs and targeted therapeutics for % any AE, % grade ≥3 AE, % discontinued due to AE, and % treatment related deaths are listed in Table 2. The median % any AE was 64% (60.7–69.3%) among the key ICI clinical trials described in this review. For the ICIs, fatigue was the most commonly observed AE. Other observed AEs specifically related to ICI therapy were asthenia, infusion-related reactions, diarrhea, anorexia, peripheral edema, pruritus and rash. Severe AEs of grade ≥3 were fatigue, anemia, hepatitis, increased lipase and amylase, diarrhea, and asthenia. Grade 5 treatment-related pneumonitis that resulted in death, occurred in patients treated with durvalumab, nivolumab, and avelumab [31,33,46]. Nivolumab caused a death due to respiratory failure. Durvalumab caused a death due to autoimmune hepatitis. Pembrolizumab caused death due to sepsis and myositis [28].

For EV, SG, and erdafitinib, the most common treatment-related AEs was also fatigue. Other AEs common for the targeted therapeutics were diarrhea, nausea, any peripheral neuropathy, neutropenia, alopecia, any rash, decreased appetite and dysgeusia, hyperphosphatemia, and stomatitis. Severe AEs of grade ≥3 were neutropenia, anemia, hypophosphatemia, fatigue, leukopenia, hyponatremia, stomatitis, and asthenia. The most common reason for treatment discontinuation were retinal pigment epithelium, hand-foot syndrome, dry mouth, and skin or nail events [35,36,72]. Notably, thus far there have been no treatment-related deaths. However, a note of caution is that there were increased proportions of patients that discontinued treatment relative to ICIs. The median % discontinued due to AE was 6% (1.6–9.2%). In contrast, EV, SG and erdafitinib had % discontinued due to AE of 1.5–2.2-fold higher. In addition, the % grade ≥3 AE category the targeted therapeutics was increased by a factor of 2.6–3.4. Although the targeted therapeutics appear to be sufficiently tolerated or AEs are appropriately managed, they are more toxic than ICIs and thus the safety of patients should be closely monitored.



Second-line setting. First-line setting. NR = not reported. Tagawa et al. [66], reports that the AE profile for sacituzumab govitecan in patients with bladder cancer was consistentwith prior reports for breast cancer. Thus, the % of any AE and % grade ≥3 AE are from the reported clinical trial of sacituzumab govitecan in patients with metastatic triple-negativebreast cancer, which was the basis for its approval for this cancer type [72], should be taken with caution.

### **6. Health Economic Factors**

The innovative therapeutic approach brought by ICIs has undoubtedly ushered a new paradigm for treating patients with metastatic bladder cancer. ICIs have provided physicians the ability to control tumor growth, extend survival, and can be administered with a better safety profile compared to traditional chemotherapy. This is a major advancement for a disease that afflicts patients that are typically elderly and are frail or have co-morbidities, and cannot tolerate harsh chemotherapy. However, the exorbitant cost of these ICIs combined with the latest follow-up results (Table 1) demonstrating either less than hoped for patient responses or failure to meet endpoints in critical phase III trials that were a condition of their accelerated approval, their high economic burden and cost-effectiveness is now widely debated [74–76].

#### *6.1. Drug Costs*

Table 3 shows the current price (\$US)/mg of ICI, a typical dose, and the price/dose. The current attitude is that these drugs cost too much [74]. Renner et al., has pointed out that ICIs may reduce toxicity but they are financially toxic and high costs limit their access in many countries inside and outside the U.S. [77]. A potential drawback for targeted therapeutics, the ADCs EV and SG and, the small molecule erdafitinib are more expensive than the listed ICIs. Unlike ICIs, ADCs are composed of three key components (antibody, chemical crosslinker, and cytotoxic payload). ADC construction involves chemical conjugation steps that can be complicated and make production and purification difficult.


**Table 3.** Pricing for ICIs and targeted therapeutics.

CE = Cost-effective. <sup>1</sup> Prices in U.S. currency. <sup>2</sup> At a \$100,000 willing-to-pay threshold. <sup>3</sup> Calculated for 90 kg person. <sup>4</sup> For 'after platinum-containing chemotherapy'. <sup>5</sup> For 'cisplatin-ineligible'. <sup>6</sup> For 28-day supply.

#### *6.2. Cost-E*ff*ectiveness*

6.2.1. Pembrolizumab for Patients Who Have Progressed within 12 Months of Neoadjuvant or Adjuvant Platinum-Containing Chemotherapy Regardless of PD-L1 Expression

The National Comprehensive Cancer Network and the European Society for Medical Oncology recommend pembrolizumab in their treatment guidelines for patients who relapse after any platinum-containing chemotherapy [18,78]. This was due to the evidence from the KEYNOTE-045 trial and pembrolizumab is the only one of the five approved ICIs to demonstrate increased survival compared to standard chemotherapy after progression on platinum-containing chemotherapy.

However, there appears to be uncertainty in the cost-effectiveness of pembrolizumab. A 2018 analysis by Sarfaty et al., based off data from KEYNOTE-045 determined that for this indication relative to chemotherapy pembrolizumab did not produce the quality-adjusted life years (QALY) gains at a willingness-to-pay threshold of \$100,000 in the U.S. [79]. QALY is a measure of the incremental health improvement provided by a new treatment compared to previous treatment options. The cost-effectiveness ratio for pembrolizumab was calculated at \$122,557/QALY in the U.S. The author's did find that pembrolizumab was cost-effective in Canada, Australia, and the United Kingdom (UK), as the costs were below the \$100,000 threshold. This finding was reliant on short-term data obtained from the trial. In comparison, Slater et al. analyzed results from the KEYNOTE-045 study

at a median follow-up of >2-years [80]. The study reported that at a willingness-to-pay threshold of \$100,000, pembrolizumab is a cost-effective option (\$93,481/QALY gained) compared to chemotherapy in the U.S.

Unlike in the U.S., many countries have cost-effectiveness assessment agencies. In March 2020, the UKs National Institute for Health and Care Excellence (NICE) Evidence Review Group (ERG) recommended against the use of pembrolizumab for this indication based on cost-effectiveness estimates (Table 3) [81]. The current list price for pembrolizumab in the UK is £26.30/mg [82]. As a single treatment course is 200 mg this amounts to approximately £5260 for a single administration. Administration of pembrolizumab is recommended at 200 mg each 3 weeks until disease progression, unacceptable toxicity, or up to 24 months without disease progression. NICE projected that pembrolizumab will cost well beyond £50,000/QALY, which is NICE's limit threshold for a drug that can extend life ≥3 months in patients suffering from a disease with a life expectancy of <2-years. We believe, these results reveal that cost-effectiveness for pembrolizumab is difficult to justify.

### 6.2.2. Pembrolizumab for Patients Who Are Cisplatin-Ineligible

NICE's ERG did recommend the use of pembrolizumab for use in patients who are cisplatin-ineligible. NICE projected the cost-effectiveness of pembrolizumab for this indication will be £67,068/QALY (\$U.S. ~87,000) [83]. Merck Sharp & Dohme economic modeling projected a cost of £37,081/QALY (\$U.S. ~49,000). NICE did acknowledge that there was a level of uncertainty in the calculated cost-effectiveness projections as the data was from the KEYNOTE-052 phase II study and not a randomized phase III study. This meant that the extrapolation of OS and PFS in patients treated with pembrolizumab were compared to an independent comparator arm that received gemcitabine plus carboplatin reported by De Santis et al. [8]. In the U.S., Merck Sharp & Dohme calculated that the cost-effectiveness of pembrolizumab in this setting will be \$81,493/QALY at a willingness-to-pay threshold of \$100,000. [84] However, it is unknown if the cost-effectiveness in the UK or the U.S. of pembrolizumab has changed since the announcement that it did not reach its primary endpoints in the critical phase III KEYNOTE-361 trial [43]. Thus, we believe the cost-effectiveness for pembrolizumab is currently difficult to justify and will most likely be deemed not cost-effective.
