Efficacy of Immune Checkpoint Inhibitors in Upper Tract Urothelial Carcinomas: Current Knowledge and Future Directions

Simple Summary Upper tract urothelial carcinoma (UTUC) represents 5 to 10% of urothelial carcinoma. Their mutational profile is different as compared to bladder urothelial carcinoma (UC). While immune checkpoint inhibitors are now part of the therapeutic landscape of urothelial carcinoma, data concerning their use in UTUC patient’s treatment remain scarce. The aim of this review is to summarize the available evidence and the biological rationale of using immune checkpoint inhibitors in high-grade UTUC. We reviewed the latest molecular characterization data and proposed an insight for future therapeutic strategies based on molecular alteration profiles. Abstract Upper tract urothelial carcinoma (UTUC) represents a rare and aggressive malignancy arising from the renal pelvis or ureter. It can develop sporadically or have a hereditary origin, such as Lynch syndrome, caused by DNA mismatch repair deficiency, leading to microsatellite instability phenotype. According to molecular characterization studies, UTUC presents different mutational profiles as compared to urinary bladder urothelial carcinomas. In particular, it has been reported that UTUC harbored a higher level of FGFR3 alterations associated with a T-cell depleted immune microenvironment. The therapeutic landscape in urothelial carcinoma is rapidly evolving, with immune checkpoint inhibitors forming part of the standard of care. A greater understanding of the molecular alterations and immune microenvironment leads to the development of new treatment combinations and targeted therapy. This review summarizes the available evidence concerning the use of immune checkpoint inhibitors and the biological rationale underlying their use in high-grade UTUC.


Introduction
Urothelial carcinoma (UC) represents the fourth most common malignancy worldwide, with an urgent need for tailored approaches in the management of the metastatic disease [1]. Depending on the level of muscle invasion seen on the pathological exam, UC is divided into muscle-invasive (MI) and non-muscle invasive (NMI) disease. MIUC of the bladder represents 25% of tumors [2] as compared to 60% in upper tract urothelial carcinoma (UTUC), explaining their increased aggressiveness [3][4][5][6]. The 5-year extravesical recurrence and overall survival rates are 28% and 23% for UTUC and bladder UC, respectively [7].
While bladder origin represents 90-95% of UCs, UTUC is less common. It represents 5-10% of UCs and can arise within the renal pelvis or ureter, which are derived from a different embryologic origin as compared to the bladder [1,8].
There is a strong relationship between UC of the bladder and UTUC since approximately 50% of patients with UTUC will have urinary bladder urothelial carcinomas either at presentation or subsequently, justifying the need to perform annual cystoscopy in the follow-up of these patients [1].
For high-risk localized disease, nephroureterectomy along with peri-operative chemotherapy is the standard of care management approach [1]. In the metastatic setting, platinum-based chemotherapy regimen remains the first-line recommended treatment [1]. However, there is a growing body of evidence concerning the use of immune checkpoint inhibitors (ICI) in the treatment of urothelial carcinoma [1,9] with the approval of several compounds in the first and second-line settings of advanced UCs. However, given their rarity, patients with UTUC represent a minority of patients included in clinical trials, and there is a paucity of data concerning ICI use in this setting.
UTUC has a different behavior as compared to bladder UC [10], and while molecular alterations of urothelial bladder carcinoma have been widely studied by The Cancer Genome Atlas, data about such alterations in UTUC remain scarce [4,11,12]. However, the novel molecular insights provided by these studies led to a better understanding of this aggressive disease and provided a rationale for new therapeutic approaches. This review summarizes the available literature regarding the use of ICIs and the biological rationale underlying their use in high-grade urothelial upper tract carcinoma management.
UTUC is represented in the spectrum of Lynch syndrome, an autosomal-dominant familial cancer syndrome caused by germline mutations in the DNA mismatch repair (MMR) genes MLH1, MSH2, MSH6, or PMS2 [26][27][28]. It is particularly associated with MSH2 mutations, with such mutations found in approximately 70% of UC integrated in Lynch syndrome [26,27,29,30]. Loss of function in the MMR system, either caused by an inherited mutation or a sporadic event, results in microsatellite instability (MSI) throughout the genome [31]. MSI have been found between 3.9% to 20% in UTUC as compared to <1% in UBC [32,33]. High MSI is correlated with a better prognosis, particularly in patients younger than 71 years old with T2-T3N0M0 tumors [34]. Furthermore, high MSI is associated with a higher tumor mutational burden (TMB) and correlated to a higher response rate with ICI treatment [35,36]. Recent retrospective data in advanced high MSI UTUC patients treated with immune checkpoint inhibitors have demonstrated excellent clinical activity [28].
Using a combination of whole-exome sequencing and RNA sequencing, some studies described biological differences between UTUC and bladder UC [11]. Four molecular subgroups of UTUC were defined through RNA-sequencing by Moss et al. [4]. Comparison to the TCGA dataset has revealed that cluster 1 was similar to luminal-subtype, cluster 2 was close to basal subtype, and cluster 4 demonstrated a high frequency of upregulated immune checkpoint related genes [4]. Moreover, UTUC is found to be predominantly luminal-like tumors, and, more precisely, luminal-papillary and T-cell depleted as compared to bladder UC [11,14].
Recently, we reported the UTUC methylation profiles and identified 2 epigenetic subtypes, namely EpiC-low and EpiC-high. While the earlier one was hypomethylated, immune depleted, and enriched for FGFR3 mutations, the latter was hypermethylated, immune infiltrated, and associated with SWI/SNF genes somatic mutations [12] (Figure 1). Moreover, we identified for the first time a high rate of mutations in the ZFP36 family in almost one-quarter of UTUC. Further mechanistic studies showed that ZFP36L1 loss of function experiments in urothelial cell lines led to increased cell migration and epithelial-mesenchymal transition [12]. More recently, Fujii et al. conducted a comprehensive molecular study of 198 UTUC patients through whole-exome sequencing (WES), single nucleotide polymorphism (SNP) array, RNA sequencing, and methylation analysis [37]. They reported 5 molecular subtypes (hypermutated (5.5%), TP53/MDM2 (37.7%), RAS (15.1%), FGFR3 (35.2%) and triple-negative (6.5%)) that correlates with clinicopathological features [37]. Indeed, patients classified in TP53/MDM2 and triple-negative subgroups had the worse disease-specific survival. Moreover, similar to previous studies, they found a majority of luminal-like tumors (71.5%) [37]. They also confirmed a different mutational profile between tumors arising in ureters versus renal pelvis, suggesting a distinct carcinogenesis mechanism depending on location in the urothelium [37]. They also sequenced urinary sediment-derived DNA and found 82.2% (95% CI, 71.5−90.2%) of sensitivity and 100% (95% CI, 81.5−100%) of specificity of this approach for UTUC diagnosis [37].  [12].
Broadly, upper-tract urothelial carcinomas can be divided into two subtypes, namely EpiC-high and EpiC-low. Epic-low subtype is hypomethylated, immune-desert, and characterized by FGFR3 somatic mutations with potential efficacy of the combination of FGFR3 immunotherapy and immune checkpoint inhibitors (ICI). Conversely, EpiC-high subtype is hypermethylated, immune-inflamed, and enriched with somatic mutations of SWI/SNF genes with potential benefit for ICI.

The Immune Microenvironment of UTUC
The immune landscape of tumors is an important determinant of the host's anti- Potential target for ICI and anti-FGFR3 inhibitors Potential target for ICI Figure 1. Proposal of molecular subtypes classification of upper-tract urothelial carcinomas adapted from Su et al. [12].
Broadly, upper-tract urothelial carcinomas can be divided into two subtypes, namely EpiC-high and EpiC-low. Epic-low subtype is hypomethylated, immune-desert, and characterized by FGFR3 somatic mutations with potential efficacy of the combination of FGFR3 immunotherapy and immune checkpoint inhibitors (ICI). Conversely, EpiC-high subtype is hypermethylated, immune-inflamed, and enriched with somatic mutations of SWI/SNF genes with potential benefit for ICI.

The Immune Microenvironment of UTUC
The immune landscape of tumors is an important determinant of the host's antitumoral response and clinical outcomes [38]. According to the results of molecular profiling studies, it is well established that sporadic UTUC displays a high rate of FGFR3 mutations (40-80%) [4,5,11,13,22]. Moreover, it has also been shown that upper tract tumors with FGFR3-alterations express lower CD8 T-cell gene signatures [11,39]. In addition, FGFR3 upregulation has been shown to be important in shaping the observed T-cell depleted phenotype [11]. Indeed, genes related to IFNG pathway, such as BTS2 and IRF9, were found downregulated in T-cell depleted cluster and the use of FGFR3 inhibitors correlated with upregulation of these genes [11]. However, how FGFR3 signaling mediates IFNG related genes and a T-cell depleted microenvironment has to be determined. Recently, Rose et al. found that upregulated FGFR3 signaling significantly correlates with upregulated PPARG gene signatures [40]. They suggested it as a potential mechanism given that the upregulation of PPARG signaling appears to suppress pro-inflammatory cytokine signaling and correlates with a non-T cell-inflamed phenotype [40].
Therefore, sporadic UTUC with FGFR3 mutations might be considered immune cold tumors. Consistent with previous studies, we recently observed similar findings. In addition, we found that FGFR3 mutated UTUC were hypomethylated compared to FGFR3 wild-type tumors, suggesting crosstalk between genetic and epigenetic phenotypes of these tumors [12]. Notably, FGFR3 wild-type tumors harbored a high rate of SWI/SNF genetic tumor alterations and were associated with a higher level of tumor infiltrated lymphocytes (TILs) [12]. Further studies are needed to fully understand these observations at the mechanistic level allowing us to elucidate better how the genetic alterations of the cancer cells might shape the immune contexture. Importantly, UTUC developed in a Lynch syndrome context because of an MMR loss of function could be considered immune hot tumors [21].
Finally, correlations were observed between PD-L1 positivity on tumoral cells and the worse outcomes in UTUC patients, although PD-L1 expression rate varied between 20−25% depending on the study and the cutoff used [41,42].

Localized Disease
The gold standard treatment for high-risk disease, defined by at least one of the following criteria: hydronephrosis, tumor size ≥ 2 cm, high-grade cytology, high-grade on biopsy, multifocal disease, previous radical cystectomy for bladder cancer, and variant histology [1], the standard is radical nephro-ureterectomy (RNU) with bladder cuff excision associated with lymph node dissection, according to the European Association of Urology (EAU) guidelines [1]. Kidney-sparing surgery could be proposed for patients with low-risk disease or even in patients with high-risk tumors located in the distal ureter to reduce morbidity. This strategy could allow optimal adjuvant chemotherapy administration by preserving kidney function [1]. For locally advanced UTUC, adjuvant chemotherapy with platinum-gemcitabine combination chemotherapy is recommended.
Indeed, the POUT trial, a multi-center randomized controlled trial including 261 patients, has reported a significant improvement in disease-free survival (DFS) (hazard ratio (HR): 0.45, 95% CI 0.30-0.68; p = 0.0001) at a median follow-up of 30.3 months (IQR 18.0-47.5) with adjuvant platinum-gemcitabine combination chemotherapy beginning within 90 days following nephroureterectomy [43]. The recently updated results of this trial, with a median follow-up of 48.1 months (IQR: 36.0-60.1), confirmed the chemotherapy benefit in DFS (HR: 0.50, 95% CI: 0.34-0.75; p = 0.001) [44]. There was no detrimental longterm toxicity observed. Furthermore, no statistically significant improvement in OS was observed [44]. Unfortunately, the optimal administration of cisplatin was not feasible for some patients due to the surgical impact on renal function [45,46]. Therefore, neoadjuvant chemotherapy was proposed, with interesting results in terms of pathological downstaging and survival [47]. Moreover, there is a benefit in terms of pathological response, disease recurrence, and mortality rate compared to RNU alone, even if no randomized controlled trials have been published yet [48]. The use of ICIs in the peri-operative setting is not yet a standard of care in UTUC management [1]. Recently published data about the use of ICI in this setting are discussed below.

Metastatic Disease
The management of the metastatic disease should be multimodal, with the use of local therapy in the case of oligometastatic disease, even in the absence of a randomized controlled trial [1,49]. RNU could also be performed in the oligometastatic situation, although providing only a small benefit [50][51][52][53], or in the palliative setting to manage symptomatic disease [1,54].
In the metastatic setting, systemic chemotherapy is extrapolated from data collected in metastatic bladder UC, since few data are available for advanced UTUC [1]. Indeed, cisplatin-based regimens, such as MVAC (methotrexate, vinblastine, doxorubicin, and cisplatin), or GC (gemcitabine and cisplatin) remained standard of care in first-line. Furthermore, a recent retrospective study found that the location of the primary tumor (upper vs. lower tract) did not change the progression-free survival (PFS) and overall survival (OS) in patients treated with platinum-based chemotherapy for metastatic UC [52]. Immune checkpoint inhibitors are also part of the therapeutic armamentarium.
Recently, avelumab as first-line maintenance therapy for platinum-based chemotherapy responder or stable disease became the standard of care [55]. The JAVELIN-100 trial is a randomized controlled phase 3 trial comparing avelumab versus placebo as maintenance therapy after platinum-based chemotherapy in first-line for patients with metastatic UC. A total of 700 patients were randomized, 30.3% (n = 106/350) versus 23.1% (n = 81/350) of UTUC patients were, respectively, in the avelumab and placebo arm. Median overall survival was significantly longer in the avelumab arm, 21.4 (18.9-26.1) versus 14.3 (12.9-17.9) hazard ratio for death 0.69 (0.56-0.86) p = 0.001 [55]. Based on subgroup analysis, the second-line systemic standard of care includes ICI, such as pembrolizumab. Indeed, in the phase 3 trial KEYNOTE-045, 542 UC patients were randomly assigned between pembrolizumab or chemotherapy chosen by the investigator. There was 14.1% (n = 38/270) UTUC patients in the pembrolizumab arm versus 13.6% (n = 37/272) in the chemotherapy arm. In the total population, median overall survival was 10.3 months (95% CI, 8.0 to 11.8) versus 7.4 months (95% CI, 6.1 to 8.3) in the chemotherapy group (HR, 0.73; 95% CI, 0.59 to 0.91; p = 0.002) [56,57]. Nevertheless, based on subgroup analysis, a systemic chemotherapy regimen with a single agent such as paclitaxel, docetaxel, or vinflunine could still be proposed [58].

Immune Checkpoint Inhibition in UTUC
Given their relative rarity, there are no studies specifically focusing on UTUC. Therefore, data related to ICI efficacy are extracted from a larger cohort of patients with UC that included a small subgroup of UTUC.

Immune Checkpoint Inhibitors in the Perioperative Setting
In the adjuvant setting, ICI-based therapy has been widely tested since the role of adjuvant treatment in high-risk muscle-invasive urothelial carcinoma after radical surgery was not clear.

Immune Checkpoint Inhibitors in the Metastatic Setting
In the metastatic setting, ICI are widely used in the management of UC (Table 2). Indeed, avelumab as maintenance therapy after platinum-based chemotherapy is currently the standard of care according to the results of the JAVELIN-100 trial [55]. For cisplatin-ineligible UC patients, based on phase 2 trials IMvigor 210 and KEYNOTE 052 provided interesting results for the use of ICI in this frail population (Table 2) [66,67]. The overall response rate for UTUC patients was 39% with atezolizumab [66] and 22% with pembrolizumab in monotherapy in this setting [67]. The IMvigor 130 trial, a randomized phase 3 trial, showed significant PFS improvement of the addition of atezolizumab to platinum-based chemotherapy [68] (Table 2). Specific outcomes of UTUC patients were not assessed. However, unfortunately, atezolizumab in monotherapy failed to improve overall survival compared to chemotherapy in pretreated metastatic UC [70]. In the same manner, the addition of pembrolizumab to first-line platinum-based chemotherapy was not associated with a survival benefit compared to chemotherapy alone (Table 2) [69].
In the second or later line, atezolizumab, durvalumab, avelumab, nivolumab, and pembrolizumab have been demonstrated safe and efficient in platinum pretreated UC population. However, UTUC patients' data remained scarce (Table 2).

Perspectives in UTUC Management
According to the recent advances in the molecular characterization of UTUC, there is a rationale to develop new treatment combinations. Indeed, given the high prevalence of FGFR3 mutations and their association with a T-cell depleted phenotype in UTUC, there is a rationale for combining ICI with FGFR3 inhibitors (Figure 1). Erdafitinib, a pan-FGFR inhibitor, is now approved based on the results of the phase 2 trial in metastatic bladder cancer, with a 40% of response rate in patients with FGFR actionable alterations [71]. Moreover, Ding et al. reported the case of a 67 years old metastatic, chemo-refractory UTUC's patient having a dramatic response to pembrolizumab in association with erdafitinib [72]. However, reliable response biomarkers are still needed to improve precision medicine in urothelial carcinoma. The ongoing trials assessing immune checkpoint inhibitors-based combinations therapies in UTUC metastatic setting are reported in Table 3. They often include backbone ICI in combination with chemotherapy, antibody-drug conjugates, and tyrosine kinase inhibitors.

Conclusions
ICI inhibitors are now widely used in daily practice to treat urothelial carcinoma patients. Based on the recent advancement in the comprehension of the molecular biology of UTUC and the differences between bladder UC and UTUC, further studies focused on UTUC patients are needed to personalize the therapeutic approach and find new treatment combinations. Conflicts of Interest: J.T. declares no conflict of interest related to this manuscript, N.M.C. reports support for research travel from Pfizer, Janssen and Ipsen, and consulting fees for BMS, Pfizer, Sanofi and Bayer, O.A. declares no conflict of interest related to this manuscript, H.L. declares no conflict of interest related to this manuscript, N.M.T. reports grants from Bristol Myers Squibb during the conduct of the study; grants and personal fees from Exelixisa and Nektar Therapeutics; grants from Calithera Bioscience, Takeda, Arrowhead Pharmaceuticals, and Epizyme, Inc.; personal fees from Bristol Myers Squibb, Pfizer, Lilly Oncology, Neolukin Therapeutics, Eisai, Novartis, Oncorena, Ono Pharmaceutical, Surface Oncology, and Ipsen outside the submitted work, P.B. reports other from BMS (consulting, travel); other from MSD (consulting, travel), Pfizer (consulting, travel), Ipsen (consulting, travel), Janssen Cilag (consulting, travel), AstraZeneca (consulting), Amgen (consulting, travel), Novartis (consulting), Eusapharma (consulting), Roche (consulting, travel), Esai (consulting, travel), Merck (consulting, travel) and Astellas (consulting, travel) outside the submitted work, G.G.M. reports consulting fees from Pfizer, Ipsen, BMS, Astellas; received travel support from Pfizer, Ipsen, BMS, MSD and Astellas outside of the submitted work.