BACKGROUND:Severe imatinib toxicity is a rare, but relevant event in GIST patients. Other approved treatments were not evaluated within a first-line setting and exhibit severe adverse events that limit QoL. Nilotinib was evaluated within this setting in GIST with a safety profile comparable to imatinib and effectiveness within KIT-exon-11-mutant GIST subgroup (ENESTg1). MATERIALS:, Patients and Methods: This is a retrospective case series reporting the outcome and side effects of 20 patients treated with nilotinib following severe imatinib toxicity. Nilotinib efficacy was tested in GIST cell lines carrying common primary and resistance mutations in KIT. Inhibitory profile of nilotinib was characterized using in silico modeling. RESULTS:In our retrospective analysis of 1263 patients, 7.3% of patients discontinued imatinib due to toxicity. In 20 patients who received nilotinib, reasons were skin (n = 11) and liver (n = 4) toxicity, followed by cardiac (n = 2) and gastrointestinal toxicities (n = 2). No cross-toxicities were observed with nilotinib treatment. Nilotinib was effective after imatinib intolerance. A secondary KIT Exon 9 mutation was found in a progressing lesion. In vitro viability assays showed effectiveness in KIT-Exon-11-mutant cell lines, but not in Exon 9 or imatinib-resistant mutations in doses deemed clinically achievable. In silico modeling revealed steric hindrance by Exon 9 or imatinib resistance mutations. CONCLUSION:Nilotinib exhibits no cross toxicity with imatinib and represents an important alternative in imatinib-sensitive, KIT-exon-11-mutant GIST given its excellent toxicity profile. There is no biological rationale for the use of nilotinib in imatinib-resistant GIST with secondary KIT mutations.
BACKGROUND:Reliable biomarkers for risk stratification and treatment monitoring in recurrent/metastatic nasopharyngeal carcinoma (R/M NPC) treated with first-line chemo-immunotherapy remain limited. This study evaluated the prognostic value of baseline computed tomography (CT)-derived body composition parameters and their early dynamic changes. METHODS:In this single-center retrospective cohort study, patients with R/M NPC receiving first-line chemo-immunotherapy were included. The primary endpoint was progression-free survival (PFS), and secondary endpoints included overall survival (OS), objective response rate (ORR), and disease control rate (DCR). Associations between body composition and outcomes were analyzed. RESULTS:A total of 199 patients were included, and 99 with post-treatment CT after two cycles were further analyzed for dynamic changes. Higher baseline skeletal muscle area (SMA) was independently associated with longer PFS (HR = 0.56, 95% CI: 0.37-0.85, p = 0.006) and OS (HR = 0.46, 95% CI: 0.23-0.93, p = 0.03), and with higher ORR (82.8% vs 70.0%, p = 0.033) and DCR (98.0% vs 91.0%, p = 0.031). During early treatment, a ≤ 1.3% decrease or an increase in SMA was independently associated with longer PFS (HR = 0.53, p = 0.045) and OS (HR = 0.27, p = 0.018). A greater decline in visceral adipose tissue density (ΔVATD < -0.9%) was also associated with more favorable PFS and OS, whereas no significant difference in ORR was observed. CONCLUSION:Baseline SMA and early changes in SMA and VATD have prognostic value in R/M NPC treated with first-line chemo-immunotherapy. CT-based body composition assessment may serve as a practical imaging biomarker for risk re-stratification and individualized management. IMPLICATIONS FOR PRACTICE:CT-based body composition assessment may provide a practical approach for individualized management of patients with recurrent/metastatic nasopharyngeal carcinoma receiving chemo-immunotherapy. Incorporating body composition assessment into routine CT imaging analysis may further improve risk stratification and facilitate early identification of patients with poor prognosis. For patients with low skeletal muscle reserve or significant early muscle loss during treatment, timely nutritional assessment, exercise support, and closer clinical monitoring may be considered to optimize supportive care and treatment management.
Abstract Background Immune checkpoint inhibitors improve outcomes in clear cell renal cell carcinoma (ccRCC), yet most patients do not achieve durable responses due in part to insufficient tumor antigenicity. While neoantigen vaccines have shown promising early-phase results in ccRCC, their patient-specific design limits broad applicability and poses additional challenges in low mutational burden tumors, where targetable epitopes are scarce. Human endogenous retroviruses (hERVs), which comprise ∼8% of the human genome, can become aberrantly expressed in cancer and represent a source of shared, tumor-specific antigens; however, the extent to which they expand the limited antigenic landscape of not only ccRCC, but non-ccRCC subtypes, remains largely unexplored. We hypothesized that a subset of these aberrantly expressed hERVs provide a new public source of immunogenic peptides that can be prioritized as candidates for mRNA vaccine development. Methods We performed a pan-cancer (TCGA) analysis of 3,173 putative hERV loci from the Vargiu et al (Retrovirology, 2016) reference to identify hERVs upregulated across all cancer types, with a focus on kidney cancer subtypes. Candidates were prioritized using a scoring framework that enriched for hERVs with high tumor expression and minimal expression in normal tissues. From prioritized loci, six-frame translation was used to generate candidate open reading frames (ORFs), which were filtered and integrated with polysome sequencing data from RCC cell lines to identify translation-supported regions. Candidate peptides were evaluated for HLA binding using complimentary antigen prediction pipelines (e.g., HLAthena, netMHCpan) for various HLA alleles. To assess immunogenicity, we developed a multi-epitope hERV-targeting mRNA vaccine and evaluated antigen-specific immune responses in HLA-A11 transgenic mouse models. Results TCGA analysis combined with our scoring framework identified ≥25 unique hERV vaccine candidates per RCC subtype, with substantial overlap between ccRCC and papillary RCC (pRCC) and a distinct profile in chromophobe RCC (chRCC). Within ccRCC-prioritized hERV candidates, ORFs were generated via six-frame translation, filtered for canonical structure, and integrated with polysome sequencing data from three ccRCC cell lines to identify regions with strong translational support. This reduced 294,927 candidate ORFs to 805 high-confidence ORFs from 18 unique hERV loci, including recently described HIF-2α–regulated hERVs 4818 and 5875. HLA binding prediction identified numerous high-affinity candidate peptides across all HLA alleles (e.g., HLA-A*02, HLA-A*11), with some loci yielding >80 predicted binders for HLA-A*11 alone. Reanalysis of published immunopeptidomics datasets identified multiple peptides mapping to prioritized hERV loci, supporting endogenous processing and presentation. A pilot study using a multi-epitope mRNA vaccine encoding previously described hERV-derived peptides in a prophylactic setting elicited robust tetramer-positive, antigen-specific T cell responses in HLA-A11 transgenic mice. Compared with a peptide-based counterpart, mRNA vaccination generated superior responses, with approximately threefold and twofold higher frequencies of tetramer-positive CD8+ T cells in the spleen and vaccine-draining lymph nodes, respectively. Conclusions We present an integrative framework to identify translated hERV-derived antigens across RCC, expanding on prior work in ccRCC by identifying both known and novel immunogenic candidate peptides with evidence of translation and strong predicted HLA binding. mRNA vaccination using previously described hERV-derived peptides elicited robust antigen-specific CD8+ T cell responses, outperforming peptide vaccination. Our findings also provide initial insight into the hERV landscape in pRCC and chRCC, supporting further investigation. Future work will evaluate newly predicted peptides in multi-epitope mRNA vaccines and test their anti-tumor efficacy in vivo. DOD CDMRP Funding yes
Abstract Background In the randomized phase 3 LITESPARK-005 study (NCT04195750), belzutifan significantly improved progression-free survival (PFS) and objective response rate, but not overall survival (OS), when compared with everolimus in participants with advanced clear cell renal cell carcinoma (ccRCC) previously treated with an immune checkpoint inhibitor and antiangiogenic therapy. Patient-reported outcomes (PROs) assessed using the Functional Assessment of Cancer Therapy-Kidney Cancer Symptom Index–Disease Related Symptoms (FKSI-DRS) and the EORTC Quality of Life Questionnaire Core 30 (QLQ-C30) global health status/quality of life (GHS/QoL) favored belzutifan. We conducted a post hoc analysis to evaluate associations between health-related quality of life (HRQoL) and clinical outcomes after >2 years of minimum follow-up. Methods Adults with advanced ccRCC were randomly assigned 1:1 to oral belzutifan 120 mg or everolimus 10 mg once daily until progression or unacceptable adverse events. This post hoc analysis population included all treated participants with ≥1 PRO assessment using the FKSI-DRS and QLQ-C30 GHS/QoL. Associations between HRQoL (baseline, longitudinal time-dependent changes, and landmark time points at weeks 13, 25 and 53) and PFS and OS were evaluated using Cox proportional hazards models. Clinically meaningful improvement from baseline was defined as a ≥ 3‑point increase for FKSI‑DRS and a ≥ 10‑point increase for QLQ‑C30 GHS/QoL per established thresholds. Landmark analysis excluded participants with events prior to each time point. No adjustments were made for multiplicity, and no formal hypothesis testing was performed. Results A total of 720 participants were included (belzutifan, n = 366; everolimus, n = 354), with a median study follow-up of 35.8 months (range, 26.9-49.2). Higher baseline FKSI-DRS scores were associated with modest PFS (hazard ratio [HR], 0.91; 95% CI, 0.86-0.96) and OS (HR, 0.86; 95% CI, 0.81-0.91) benefit. Higher baseline QLQ‑C30 GHS/QoL scores were not associated with PFS benefit (HR, 0.98; 95% CI, 0.94-1.03) but were associated with a small OS benefit (HR, 0.92; 95% CI, 0.88-0.96). In longitudinal analyses, improvement in FKSI-DRS was associated with lower risk of progression or death (HR, 0.80; 95% CI, 0.74-0.85) and death (HR, 0.57; 95% CI, 0.53-0.62); similar associations were observed for QLQ-C30 GHS/QoL (PFS: HR, 0.80; 95% CI, 0.74-0.85; OS: HR, 0.64; 95% CI, 0.60-0.69). In landmark analyses, participants with stable or improved HRQoL at weeks 13, 25, and 53 had numerically longer PFS and OS compared with those with deterioration, although estimates were imprecise at later time points (Table). Conclusions In this exploratory post hoc analysis of LITESPARK-005, higher baseline HRQoL and improvements during treatment were associated with improved PFS and OS in advanced ccRCC. These hypothesis-generating findings support HRQoL as a potential prognostic marker; however, given the pooled treatment arms and potential for time-dependent confounding and reverse causation, results should be interpreted cautiously.
Abstract Background Clear cell renal cell carcinoma (ccRCC) with sarcomatoid differentiation (sRCC) is associated with poor survival. Recent studies have shown downregulation of hypoxia-related pathways in sRCC (Motzer et al., Cancer Cell, 2020; El Zarif, Semaan et al., Cell Reports, 2024). In this study, we sought to compare HIF2α expression levels in ccRCC, with and without sarcomatoid differentiation and investigate clinical outcomes of patients (pts) with sRCC treated with HIF2α inhibitors (HIF2αi). Methods To assess HIF2α gene expression in sRCC, RNA-seq data was collected from 2 clinical trials: JAVELIN Renal 101 (JR101), and IMmotion151 (IM151), as well as publicly available data from The Cancer Genome Atlas (TCGA). Mean HIF2α expression levels were compared between ccRCC with and without SD using Wilcoxon Rank-Sum test. To evaluate the clinical outcomes of pts with sRCC treated with HIF2αi, we also collected clinical data from pts with ccRCC who were treated with Belzutifan between January 2018 and December 2025 at the Dana-Farber Cancer Institute. Progression-free survival (PFS) and overall survival (OS) were analyzed using log-rank test and multivariable Cox regression model accounting for age, IMDC risk group, line of therapy, and type of regimen (HIF2αi; HIF2αi + Vascular endothelial growth factor-targeted therapy (VEGF-TT)). In addition, we performed a similar survival analysis on patients who were also treated with immunotherapy-based regimens from the same cohort. Results RNA-seq data from 2,075 pts with renal cell carcinoma with a clear cell component from TCGA (48 sRCC; 493 ccRCC), JR101 (97 sRCC; 639 ccRCC) and IM151 (110 sRCC; 688 ccRCC) were included. Expression of EPAS1, which encodes for HIF2α, was significantly downregulated in pts with sRCC in TCGA (log(TPM) mean: 6.63 vs 9.95 in ccRCC ; p < 0.001), JR101 (8.71 vs 8.93 in ccRCC ; p = 0.04) and IM151 (8.05 vs 8.69 in ccRCC ; p < 0.001). For the clinical cohort, 140 pts treated with Belzutifan were included with a median follow up of 13 months. In multivariable analysis, sRCC was associated with worse PFS (HRadjusted: 2.04; 95% CI: 1.22–3.42; p = 0.006) and worse OS (HRadjusted: 2.56; 95% CI: 1.35–4.8; p = 0.004) (Figure). In parallel, from the same cohort, 120 patients were also treated with immunotherapy; 16% (19/120) had sRCC. When treated with immunotherapy-based regimens, sRCC and ccRCC patients had comparable PFS (HRadjusted: 1.13; 95% CI: 0.63–2.01; p = 0.68) and comparable OS (HRadjusted: 1.02; 95% CI: 0.99–1.05; p = 0.26). Conclusions This study suggests that the efficacy of HIF2αi may be limited in sRCC, possibly due to decreased dependency on HIF2α signaling. Our findings highlight the need for novel therapeutic strategies in sRCC.
Abstract Background Nivolumab plus ipilimumab (NIVO+IPI) combination therapy is a first-line standard in metastatic renal cell carcinoma (mRCC). Despite durable responses in some patients, over 20% develop primary progressive disease (PPD) at first restaging imaging. In this study, we evaluated clinical factors associated with PPD and subsequent outcomes. Methods This retrospective real-world study analyzed patients with mRCC who received first-line NIVO+IPI, within the International Metastatic Renal Cell Carcinoma Database Consortium (IMDC). Response to treatment was assessed by physician evaluation according to Response Evaluation Criteria in Solid Tumors (RECIST), version 1.1. Multivariable logistic regression was used to identify factors associated with PPD. Among patients with PPD, Cox proportional hazards models were used for overall survival (OS) and time to next line therapy or death (TNTD), and Fine–Gray competing risks models for mortality without next-line therapy (MWNT). Results 1,601 patients with mRCC treated with first-line NIVO+IPI were included, of whom 487 (30.4%) experienced PP. In multivariable logistic regression analysis, liver metastases (odds ratio 1.67, 95%CI 1.24-2.26), Karnofsky performance status (KPS) <80% (1.54 (1.12-2.12)), low hemoglobin (1.35 (1.04-1.74)), and elevated neutrophils (1.50 (1.09-2.06)) were independently associated with higher odds of PPD, while prior nephrectomy was associated with lower odds (0.69 (0.54-0.88)). Among patients with PPD, the median time to end of first-line therapy (TEFL) was 2.1 months. The median TNTD was 3.6 months, and the median OS was 13.9 months. At 6 months, the cumulative incidence of initiating next-line therapy was 59%, while 16% died without receiving subsequent therapy. In multivariable analyses among patients with PPD, liver metastases, treatment initiation within 1 year of diagnosis, poor performance status, and elevated neutrophils were associated with shorter TNTD and worse OS, while low KPS, elevated neutrophils, and low hemoglobin were associated with increased MWNT, and prior nephrectomy remained protective (Table 1). Conclusions Primary progressive disease to first-line NIVO+IPI occurs in around 30% of patients with mRCC and is associated with poor clinical outcomes. This real-world analysis identified baseline factors at time of initiation of treatment reflecting aggressive disease biology. These findings emphasize the importance of early risk stratification to identify patients unlikely to benefit from NIVO+IPI, who may benefit from an alternative first-line approach or an early change in treatment.
Abstract Background MiT/TFE-rearranged, translocation renal cell carcinoma (tRCC) is characterized by gene fusions involving TFE3 or TFEB, and is a rare, aggressive non-ccRCC, often occurring in children and with few biomarkers or targeted therapy options. TFE3 is the most commonly translocated gene with ASPSCR1, PRCC and SFPQ as fusion partners. GPNMB (Glycoprotein non-metastatic melanoma protein B) is a canonical transcriptional target of MiT/TFE proteins expressed at low levels in normal tissues but upregulated in numerous MiT/TFE-driven tumors (e.g. tRCC), where it is associated with poor prognosis. GPNMB can also be cleaved at its ECD by metalloproteinases and secreted from the surface of cancer cells. Soluble GPNMB can thus be detected in the serum of cancer patients and function as a surrogate biomarker. CD44, the primary receptor for GPNMB, is a tumor-associated antigen linked to cancer invasion, metastases and stemness, and is associated with poor prognosis in many cancers. The aim of the current study was to characterize the biology of GPNMB and CD44 in pre-clinical models of tRCC Methods We used the following human cell line models: As in vitro pre-clinical companion models we analyzed cells with stable doxycycline-inducible expression of WT-TFE3, PRCC-TFE3, SFPQ-TFE3 and NONO-TFE3 fusions using a) the Flp-In T-RExTM system in HEK293 cells, b) rtTA3 (Tet-on) system in HK2 proximal tubular epithelial cells and c) patient derived UOK cells with TFE3-fusion transgenes, comparing them to UOK cells from ccRCC patients. We also generated a novel transgenic knock-in mouse expressing the SFPQ-TFE3 fusion. We crossed SFPQ-TFE3LSL knock-in mice to 1) KSP-Cadherin 16 Cre mice that express Cre recombinase in the distal tubular epithelial cells and collecting ducts of the kidney. 2) Tamoxifen-inducible, Pax8- ERTCre mice diffusely expressing Cre recombinase following kidney development. We also examined tumorigenesis in the previously described PRCC-TFE3LSL; Ksp-Cre mice (W. Marston Linehan, NIH), and in an ASPSCR1-TFE3 PDX mouse model and a LN metastatic lesion from a patient with the ASPSCR1-TFE3 fusion (Dr. John Copland, Mayo Clinic). We are also currently examining blood serum, EDTA plasma and urine samples from non-metastatic and metastatic ccRCC and tRCC cases, obtained courtesy of the Kidney Cancer Program at Johns Hopkins. Results To further characterize GPNMB functionality and value as a cell-surface therapeutic target in tRCC, we engineered cell lines with genomic deletion of GPNMB via CRSIPR-Cas9 editing. Deletion of GPNMB in PRCC-TFE3 cell lines [UOK120/UOK124] significantly decreased clonogenic growth in 2D, spheroid size and viability and tumor xenograft growth in NSG mice and was associated with a decrease in phosphorylation of mTORC1 substrates [p-P70S6K, p-4EBP1]. We then examined expression of CD44, the primary receptor for GPNMB, and a tumor-associated antigen associated with poor prognosis in many cancers. Expression of CD44 and its ligand SPP1/OPN, was significantly increased in bulk RNA-Seq data from multiple transgenic models of SFPQ-TFE3, PRCC-TFE3 and ASPSCR1-TFE3, in human tRCC cases compared to normal kidney, and in SFPQ-TFE3/ PRCC-TFE3 transgenic kidney tumors and an ASPSCR-TFE3 PDX model, by immunoblotting and IHC, with increased membrane localization. shRNA-mediated depletion of CD44 profoundly and specifically decreased clonogenicity of multiple TFE3-fusion lines, with no effect seen in ccRCC lines. Conclusions In conclusion, GPNMB regulates the growth of tRCC cells, potentially via an autocrine mechanism involving its receptor CD44, and targeting GPNMB-CD44 signaling may be of therapeutic benefit in tRCC. DOD CDMRP Funding yes
Abstract Background Loss of VHL, the primary driver of clear cell renal cell carcinoma (ccRCC), promotes aerobic glycolysis and suppresses oxidative phosphorylation (OxPhos). In contrast, non-clear cell RCC (non-ccRCC) comprises heterogeneous tumors lacking shared nuclear drivers. Mitochondrially-encoded electron transport chain (ETC) genes encoding for complex I, required for OxPhos, have been recently reported to be recurrently mutated across cancers. Here, we show that mitochondrial DNA (mtDNA) mutations in non-ccRCC phenocopy VHL loss, converging on a shared metabolic program to drive a stem-like transcriptional state. Methods mtDNA mutations were called in samples with sufficient coverage ( > =5 reads across >90% of the mitochondrial genome) in TCGA (N = 3,265; N = 324 RCC) and the institutional MSK-IMPACT cohort (N = 22,252; N = 568 RCC). Isogenic cell lines (ACHN, 786O, HEK293T) harboring pathogenic complex I mutations (m.11696G>A) were generated using DddA-derived cytidine base editors (ddCBEs) at different levels of heteroplasmy. Metabolic phenotyping was performed using Gas Chromatography (GC/MS) and Liquid Chromatography–Mass Spectrometry (LC/MS) as well as Seahorse and YSI assays. Single cell multi-omics was performed with DOGMA-seq with mtDNA variant calling performed using mgatk. Results Across all cancer types in TCGA, RCC tumors were among the most enriched in ETC truncating mutations. These mutations most commonly affected mitochondrial components of complex I and were enriched to high levels of heteroplasmy. VHL-driven ccRCC was relatively depleted in these mutations compared to other subtypes (ccRCC: 8.6%, chRCC: 20.0%, pRCC: 34.0%) (Figure 1a). These results were independently replicated in the MSK-IMPACT cohort (data not shown). ACHNm.11696G>A heteroplasmic cells displayed decreased oxygen consumption rates (Figure 1b), increased glucose consumption and lactate production (Figure 1c), phenocopying the effect of VHL KO (data not shown). These metabolic changes led to an increase in the 2-hydroxyglutarate/alpha-ketoglutarate (2HG/aKG) ratio with increasing heteroplasmy (Figure 1d), complex I pharmacologic inhibition (Figure 1e), and VHL KO (Figure 1f). This increase in 2HG was entirely accounted for by the L-enantiomer of 2HG (Figure 1g) and was consistently found in RCC tumors compared to normal kidney (Figure 1h), with the extent of elevation being most pronounced in tumors with sarcomatoid/rhabdoid features (Figure 1i). To evaluate the effect of heteroplasmy on transcriptional state, we performed DOGMA-seq on ACHNm.11696G>A with 76% bulk heteroplasmy (Figure 1j); the wide single-cell distribution of heteroplasmy (Figure 1k) allowed us to evaluate the correlation of heteroplasmy with transcriptional state at the single-cell level. We consistently found that heteroplasmy associated with an increase in OCT4, NANOG, and SOX2 target genes consistent with a stem-like signature (Figure 1l). Conclusions mtDNA mutations are very common in non-ccRCC and phenocopy the metabolic effects of VHL loss in ccRCC. Both mtDNA mutations and VHL loss lead to an increase in the 2HG/aKG ratio, which has been shown to lead to a block in differentiation via the inhibition of aKG-dependent dioxygenases (PMIDs: 33092942; 26212717; 31534224). Consistently, we find that increasing heteroplasmy associates with a stem-like signature. These findings nominate mtDNA mutations as a shared driver across non-ccRCC subtypes. DOD CDMRP Funding yes
Abstract Background Tivozanib is a selective vascular endothelial growth factor receptor (VEGFR) tyrosine kinase inhibitor (TKI) approved refractory in metastatic renal cell carcinoma (mRCC). However, there is limited data on the real-world efficacy of this agent. Furthermore, clinicogenomic factors associated with response to tivozanib are unknown. Methods We performed a multi-center retrospective study of patients with mRCC treated with tivozanib. Clinical information, disease characteristics, and clinicogenomic data were recorded using electronic medical records. Data were used to analyze objective response rates (ORR), disease control rate (DCR), progression-free survival (PFS), and overall survival (OS). Comparisons between responders and non-responders were performed using Fisher’s exact test. Genomic analysis was performed on patients with available next generation sequencing data (tissue and liquid). Results Among 123 patients with mRCC treated with tivozanib, 58 had adequate follow up data and were included in this analysis. Among this cohort, 69% were male, and 19% were Hispanic. The median age at the start of tivozanib was 63 years (range 29-85). Clear cell histology was the most common (96%), with non-clear cell histology representing 4% of patients. Sarcomatoid features were present in 19% of patients. At the start of therapy, the most common metastatic sites included lungs (79%), lymph nodes (40%), liver (34%), and bone (31%). Most patients (93%) previously had a nephrectomy, with 25% receiving prior adjuvant therapy. The median number of lines of therapy in a metastatic setting prior to tivozanib was 3 (range: 1-8). Among the 53 patients with available imaging, ORR was 17% and DCR was 51%. The median OS was 15.1 months (95% CI, 12.1-NE). The median PFS was 3.7 months (95% CI, 3.0-4.9). Treatment interruption and dose reduction occurred in 16% and 28% of patients, respectively. No differences in outcomes were observed between age groups, sites of metastatic disease, and prior therapies. Frequent genomic alterations included VHL (38%), PBRM1 (24%), SETD2 (21%), and TP53 (19%). Mutations in CDKN2A, CDKN2B, and NF1 had numerically response rates although not statistically significant. Conclusions In this notably diverse cohort (19% Hispanic) of heavily pretreated patients with mRCC and frequent rates of visceral disease, tivozanib demonstrated clinical efficacy in this population. Updated adverse events revealed no new safety signals. There were no significant associations between clinicogenomic factors and outcomes, underscoring the need to identify novel biomarkers for mRCC.
Abstract Background Native American (NA) populations in Oklahoma experience disproportionately high incidence (43.7 vs 20.7 per 100,000) and mortality (8.6 vs 4.7 per 100,000) from renal cell carcinoma (RCC) compared with Non-Hispanic White (NHW) populations. However, population-based datasets lack patient-level clinical detail, limiting understanding of outcomes both overall and at academic cancer centers. We characterized disease presentation, treatment patterns, and overall survival (OS) among NA RCC patients treated at Stephenson Cancer Center (SCC), stratified by disease setting, and compared stage-matched outcomes with SEER (Surveillance, Epidemiology, and End Results Program) and Oklahoma registry benchmarks. Methods We conducted a retrospective cohort study of NA patients with RCC treated at SCC between 2003 and 2024. Clinical and demographic data were obtained via EMR review. The cohort was stratified into non-metastatic (Stage I–III) and metastatic (Stage IV) groups, with Stage II/III designated as a high-risk localized subgroup. American Joint Committee on Cancer (AJCC) staging system was used to categorize stage at diagnosis. Population-level benchmarks were obtained from SEER and the Oklahoma Central Cancer Registry (OK2Cancer). OS was analyzed using Kaplan–Meier methods with log-rank testing. Stage-stratified comparisons (localized, regional, distant) were performed against SEER NA benchmarks, and a Cox proportional hazards model, adjusting for age and stage, was used to compare OS between the SCC and SEER cohorts. Results A total of 159 NA RCC cases were included, to our knowledge, representing one of the largest single-institution NA RCC cohorts with patient-level clinical data reported to date. The cohort included 136 non-metastatic (Stage I–III), 13 metastatic (Stage IV), and 10 unstaged patients. Median age at diagnosis was 60 years (range 29–83), with a male preponderance (62.3%) and predominantly clear-cell histology (94.0%). Compared with SEER-reported NA patients, SCC patients were diagnosed at a significantly younger age (60 vs 65 years, p < 0.0001). Among non-metastatic patients, 40.4% had high-risk localized disease (Stage II/III, n = 55). Relative to SEER benchmarks, the SCC cohort had a significantly lower proportion of Stage IV disease at diagnosis (8.2% vs 28.4%, p < 0.0001) and a higher proportion of stage III disease (30.8% vs 9.9% SEER NA). Nephrectomy was performed in 91.8% of patients, and 12% of patients received systemic therapy. Median recurrence-free survival among non-metastatic patients was 162.3 months (5-year RFS: 86.1%).On stage-stratified analysis, SCC NA patients demonstrated significantly improved OS compared with SEER NA benchmarks across all disease settings: localized (log-rank p < 0.0001), regional (log-rank p < 0.0001), and distant (log-rank p < 0.0001). After adjusting for age and stage, SCC patients had a 90% lower hazard of death compared with SEER NA patients (HR 0.100, p < 0.0001). Median travel distance to SCC was 78 miles (range 5.5-255 miles). Conclusions Despite a high burden of locally advanced disease at diagnosis, NA patients with RCC treated at SCC, an NCI-designated tertiary cancer center, demonstrated significantly improved OS compared with population-based benchmarks across all disease stages, even after adjusting for age and stage. Earlier stage diagnosis, high utilization of definitive surgical therapy, and access to specialized multidisciplinary care may contribute to these favorable outcomes, though the retrospective design and potential for racial misclassification in SEER data warrant caution in interpretation. In the context of persistent statewide disparities in RCC incidence and mortality among NA populations, these findings highlight the importance of expanding access to specialized cancer care and reducing geographic barriers to improve outcomes for this underserved population.
Abstract Background Belzutifan plus lenvatinib significantly improved progression-free survival and objective response rate versus cabozantinib, and had manageable safety, in participants with advanced renal cell carcinoma (RCC) that progressed on or after anti-programmed cell death protein 1 and anti-programmed cell death ligand 1 (anti–PD-[L]1) therapy in the phase 3 LITESPARK-011 study (NCT04586231). We report health-related quality of life (HRQoL) outcomes from LITESPARK-011. Methods Participants ≥18 years of age with advanced RCC that progressed on or after anti–PD-(L)1 therapy as the immediate prior treatment were randomly assigned 1:1 to receive belzutifan 120 mg plus lenvatinib 20 mg by mouth once daily or cabozantinib 60 mg by mouth once daily. HRQoL was evaluated in participants who received ≥1 dose of study treatment and completed ≥1 patient-reported outcome (PRO). Prespecified exploratory end points included time to deterioration (TTD) and least squares mean (LSM) change from baseline in Functional Assessment of Cancer Therapy-Kidney Cancer Symptom Index-Disease Related Symptoms (FKSI-DRS), and European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-Core 30 (EORTC QLQ-C30) global health status/quality of life (GHS/QoL), physical functioning, and role functioning subscales. Questionnaires were completed electronically on day 1 of week 1, 3, 5, and 9, every 4 weeks thereafter, at treatment discontinuation, and 30 days after last dose. The analysis timepoint was week 45 (last timepoint where completion and compliance rates of ∼60% and ≥80%, respectively, were observed). PROs were not formally statistically tested. Results The PRO analysis population included 731 participants (n = 365, belzutifan plus lenvatinib; n = 366, cabozantinib). At data cutoff (Apr 9, 2025), median study follow-up was 29.0 months (range, 19.3-49.2). Completion rates for FKSI-DRS and QLQ-C30 were >85% at baseline (compliance >86%) and >55% at week 45 (compliance >89%) in each arm. TTD and LSM change from baseline in FKSI-DRS, QLQ-C30 GHS/QoL, physical functioning, and role functioning scores were similar between arms (Table). Conclusions Belzutifan plus lenvatinib had similar TTD for FKSI-DRS, and GHS/QoL, physical functioning, and role functioning versus cabozantinib. LSM changes in HRQoL and disease-specific symptoms from baseline to week 45 were also similar for both arms. Combined with efficacy and safety data, results support belzutifan plus lenvatinib as a new treatment option for advanced RCC that progressed after anti–PD-(L)1 therapy. 2026 American Society of Clinical Oncology, Inc. Reused with permission. This abstract was accepted and previously presented at the 2026 ASCO Annual Meeting. All rights reserved.
Abstract Background Clear cell renal cell carcinoma (ccRCC) is the most common subtype of kidney cancer and remains a major clinical challenge. Although treatment options have expanded to include VEGF pathway inhibitors, immune checkpoint inhibitors (ICIs), and more recently a HIF2 inhibitor, many patients with metastatic disease eventually relapse, and overall survival rates remain below 50%. While ccRCC is considered an immunogenic tumor, the antigenic targets driving effective immune responses remain poorly understood. A hallmark of ccRCC is loss of the VHL gene, which leads to stabilization of hypoxia-inducible factor 2 (HIF2), a transcription factor that promotes tumor growth and survival. A key insight came from a patient who achieved complete remission following allogeneic stem cell transplantation (allo-SCT). Analysis revealed donor-derived T cells that recognized a peptide derived from a HIF2-induced endogenous retrovirus (ERV), specifically ERVE-4. This finding provided the first evidence that HIF2-driven ERV expression in ccRCC can generate tumor-specific epitopes capable of eliciting strong antitumor immune responses, suggesting a previously underappreciated source of immunogenic targets. Supporting this concept, clinical studies have shown that ERV expression correlates positively with responses to immunotherapy, further indicating that ERVs may serve as tumor-specific antigens in ccRCC. Building on this rationale, we used a multi-omics approach to systematically identify HIF-regulated ERVs. Methods To improve the accuracy of ERV detection and quantification, we utilized a comprehensive ERV database developed by Dr. Bradley Bernstein’s laboratory and performed stranded RNA sequencing to enable precise transcript identification. In parallel, targeted long-read DNA sequencing was used to resolve locus-specific ERV sequence variants, which are critical for accurate open reading frame (ORF) prediction and for constructing customized databases for mass spectrometry (MS) analysis. To identify peptides presented on ccRCC cells, we performed HLA immunoprecipitation using both endogenous HLA and engineered cell lines expressing tagged alleles, followed by immunopeptidomic profiling. Specifically, ccRCC cell lines were engineered to overexpress tagged common HLA alleles such as HLA-A*02:01, HLA-A*03:01 and HLA-B*07:02 to enable detailed characterization of HIF-regulated ERV-derived peptides presented by these alleles. Candidate ERV-derived peptides were prioritized based on known immunogenicity and recurrence across datasets. To evaluate T cell recognition, we performed ex vivo priming and expansion of peripheral blood mononuclear cells (PBMCs) from healthy donors, followed by single cell TCR sequencing, which enabled identification and characterization of T cell receptors (TCRs) specific for ERV-derived antigens. Results Integration of the updated ERV database with stranded RNA-seq substantially improved detection sensitivity, increasing the number of identified HIF-regulated ERVs from approximately 100 to nearly 300. Targeted long-read sequencing revealed numerous sequence variations at ERV loci relative to the reference genome, allowing more accurate ORF prediction and incorporation of sample-specific sequences into MS search databases. Immunopeptidomic profiling of engineered ccRCC cell lines expressing tagged common HLA alleles enabled the generation of a reference library of ERV-derived peptides presented across multiple prevalent HLA types. Importantly, we identified dozens of TCRs capable of recognizing these ERV-derived peptides, demonstrating their immunogenicity and reinforcing their relevance as tumor-specific targets. Conclusions These findings demonstrate that HIF2 drives the expression of tumor-specific ERVs in ccRCC that are actively translated and presented on the cell surface via HLA molecules across multiple common alleles. The ability of T cells to recognize these ERV-derived peptides highlights their promise as novel targets for immunotherapy. Collectively, this work provides a foundation for the development of ERV-targeted therapeutic strategies, including TCR-based therapies and cancer vaccines, to improve outcomes for patients with ccRCC.
Abstract Background Sex disparities in renal cell carcinoma (RCC) are well documented, with males exhibiting higher incidence, mortality, and more aggressive tumor features than females. While these differences have been partly attributed to risk factors such as smoking, hypertension, and obesity, emerging evidence implicates sex-specific differences in anti-tumor immunity. One proposed mechanism underlying female protection is escape from X-chromosome inactivation (XCI), which permits biallelic expression of select X-linked tumor suppressor genes (EXITS). Among these, the epigenetic regulator Kdm6a (UTX) has been implicated in modulating natural killer (NK) cell function in tumor-free mice. Notably, males exhibit increased NK cell abundance, but reduced effector function compared to females. Historically, most studies of sex bias in RCC have focused on adaptive immunity, particularly CD8+ T-cell dysfunction and exhaustion within the tumor microenvironment in males. In contrast, the contribution of innate immune mechanisms, particularly NK cell–mediated tumor surveillance, to sex-based differences in RCC progression remains poorly characterized. Methods Two ccRCC models, representing distinct clinical subtypes, were used to investigate sex-specific differences: an immune checkpoint blockade (ICB)-resistant model, Vhl⁻/⁻; p53⁻/⁻; Rb1⁻/⁻; Myc-overexpression (VpRM), and an ICB-responsive model, Vhl-mutant; Cdkn2a/2b⁻/⁻; Bap1⁻/⁻ (VCCB). Metastatic derivatives of these models were used to assess sex-dependent metastatic tropism and immune surveillance. Immune cell depletion studies targeting CD4+ T-cells, CD8+ T-cells, B-cells, NK-cells, and group 1 innate lymphoid cells were performed to define immune populations’ contributing to sex bias. Genetically engineered mouse models (GEMMs) with selective immune cell deficiencies were used to validate these findings and bone marrow chimeras to evaluate host-intrinsic determinants of sex-biased tumor progression. Results We found that sex bias in RCC models is immune dependent. In immune-deficient settings, male mice in both models exhibited smaller tumors than females; however, in immune-competent settings, female mice developed smaller tumors, recapitulating human disease. In the metastatic setting, tropism was model dependent: the immune-inflamed VCCB model showed preferential lung colonization, whereas the VpRM model preferentially seeded the liver. Notably, in the VpRM model, female mice exhibited near-complete protection from liver metastases, suggesting a highly protective or immune-surveilled female liver microenvironment. Flow cytometry revealed increased NK cells and effector ILC1s (NK1.1+, NKp46+, Eomes⁻, CD49a+, CD127⁻) in female livers. These effector ILC1s displayed higher expression of cytotoxic granzyme B and C. NK cells in females also showed increased granzyme B expression and a more mature, terminally differentiated CD11b+ phenotype, consistent with enhanced effector function. While depletion of CD4+ T-cells, CD8+ T-cells, B-cells, and ILC1s did not alter liver metastatic incidence, depletion of NK cells strikingly exacerbated liver mets in females. In Rag1⁻/⁻ mice, characterized by the absence of mature T-cells and B-cells and the enrichment of circulating and splenic NK cells, both sexes showed near-complete protection from liver metastases; this phenotype was reversed with NK cell depletion, confirming a central role for NK cells in hepatic metastasis control. Bone marrow chimera experiments where CD45.1 male and female bone marrow were transplanted into male and female CD45.2 hosts, further demonstrated that female protection is conferred by both hematopoietic and non-hematopoietic compartments, indicating that either host or NK-intrinsic female factors are sufficient to enhance immune surveillance. Conclusions Our findings identify a sex-dependent axis of innate immune surveillance, where female-intrinsic factors prime NK populations for elevated tumor immune surveillance. This highlights innate immunity as a critical, underappreciated driver of sex-based clinical outcomes in RCC. DOD CDMRP Funding yes
Abstract Background Blood biomarkers predicting benefit and toxicity from first-line (1L) immune checkpoint inhibitor (ICI) therapy in metastatic RCC (mRCC) remain limited. Eosinophils facilitate T-cell activation while neutrophils reflect systemic inflammation. We hypothesized an “inflamed eosinophilic” phenotype – characterized by low baseline neutrophil-to-eosinophil ratio (NER) with on-treatment eosinophil rise – identifies patients with favorable outcomes without increased immune-related adverse events (irAEs). Methods We retrospectively analyzed 117 mRCC patients receiving 1L ICI-based therapy at the University of Virginia (ICI/ICI 68%, ICI/tyrosine kinase inhibitor (TKI) 17%, ICI monotherapy 15%). Baseline absolute eosinophil count (AEC), NER, and neutrophil-to-lymphocyte ratio (NLR) were tested in multivariable Cox models adjusted for treatment regimen. Change in AEC (ΔAEC) was defined as percent change from baseline within 12 weeks; AEC trajectories over 12 weeks used linear mixed-effects models. Patients were classified into three predefined phenotypes: inflamed eosinophilic (IE: NER ≤40 and ΔAEC ≥10%), inflamed non-eosinophilic (INE: NER ≤40 and ΔAEC <10%), and non-inflamed (NI: NER >40). The 10% ΔAEC threshold was pre-specified based on prior nivolumab-treated mRCC data. Phenotype analysis required a calculable baseline NER (AEC >0) and at least one post-baseline AEC measurement. Associations with progression-free survival (PFS), overall survival (OS), and irAEs were assessed using Cox and logistic regression. Receiver operating characteristic (ROC) analysis was performed for 1-year OS. Results Higher baseline AEC independently predicted improved PFS (hazard ratio [HR] 0.02 per 1 K/μL, p=0.004). Elevated baseline NER was associated with inferior PFS (HR 1.008/unit, p=0.001) and OS (HR 1.006/unit, p=0.008). Optimal baseline NER cutoff for 1-year OS was 39.25 (area under the curve [AUC] 0.71). Baseline AEC trended higher in patients developing irAEs (p=0.08), whereas baseline NER and NLR were not associated with irAEs. AEC trajectories did not significantly differ by regimen and were not independently associated with survival. The IE phenotype demonstrated superior OS versus INE (HR 3.63, 95% CI 1.10-11.96, p=0.034) and NI (HR 2.88, 95% CI 1.06-7.88, p=0.039); median OS was not reached in IE vs 601 and 900 days, respectively. Median PFS was longest in IE (658 vs 460 vs 306 days). Rates of irAEs were similar across phenotypes (51.5% vs 50.0% vs 34.8%, p=0.44). Of 117 analyzed patients, 51 were excluded from phenotype analysis due to missing baseline NER or post-baseline AEC. These missing data were primarily due to baseline eosinophil counts of 0 (which prevented NER calculation) or a lack of routine post-baseline lab draws, rather than early mortality. Excluded patients did not differ significantly from included patients in baseline characteristics. Conclusions Low baseline NER with early eosinophil mobilization suggests a composite “inflamed eosinophilic” phenotype that may help identify patients with superior survival and acceptable toxicity in 1L ICI-treated mRCC. This composite approach requires validation in larger studies, but has the potential to enhance precision immunotherapy and improve patient outcomes in the expanding first-line mRCC treatment landscape.
Abstract Background In metastatic clear cell renal cell carcinoma (ccRCC), robust biomarkers of immunotherapy response remain lacking to select first-line treatment strategies. The peripheral T-cell receptor (TCR) repertoire reflects the systemic adaptive immune landscape and potential antitumor immune response. We investigated whether baseline peripheral TCR diversity was associated with outcome in patients receiving first-line dual immune checkpoint blockade (dual IO) or immunotherapy combined with a tyrosine kinase inhibitor (IO-TKI). Methods Baseline peripheral blood samples were prospectively collected (NCT04932525). Peripheral blood mononuclear cells were profiled by 5’ single-cell V(D)J sequencing (10X Genomics) and TCR clonotypes reconstructed using the scRepertoire package. Repertoire diversity was quantified using Hill q0 and Hill q1, corresponding respectively to clonotype richness, defined as the number of distinct clonotypes, and effective diversity, reflecting the relative abundance of dominant clonotypes. Associations between baseline TCR diversity and objective response according to RECIST and progression-free survival (PFS) were evaluated. Hill scores were analysed according to median-based stratification and as continuous variables. Rarefaction-based sensitivity analyses accounted for differences in sampling depth. Results Among 18 evaluable patients (9 dual IO, 9 IO-TKI), 13 (72%) achieved an objective response, including 6 (33%) in the dual IO cohort and 7 (39%) in the IO-TKI cohort. Across treatment arms, responders had numerically lower baseline TCR diversity than non-responders across all Hill metrics, although differences were not statistically significant (median Hill q0: 2310 vs 3332, p = 0.14; Hill q1: 879 vs 3248, p = 0.29). When stratified by treatment arm, this pattern was more evident in the dual IO subgroup, where responders had markedly lower diversity, with the strongest signal observed for Hill q1 (median Hill q0: 2379 vs 3332, p = 0.26; Hill q1: 786 vs 3248, p = 0.048). No difference was observed in the IO-TKI subgroup. At a median follow-up of 23.9 months, higher baseline Hill q1 values were associated with shorter PFS in the overall cohort. Median PFS was not reached in the Hill q1-low group versus 4.6 months in the Hill q1-high group (log-rank p = 0.007); continuous Hill q1 was also significantly associated with increased progression risk (HR 2.21, 95% CI 1.14-4.27, p = 0.019). Overall survival analyses showed directionally consistent trends towards improved outcomes with lower Hill q1 scores but were limited by event number. Limited sample size precluded subgroup analyses for survival. Conclusions Baseline peripheral TCR diversity may reflect clonal expansion states and impact immunotherapy outcome in metastatic ccRCC, notably in patients treated with dual IO. These findings support further investigation of peripheral TCR profiling as a non-invasive biomarker of checkpoint inhibitor sensitivity in metastatic ccRCC. DOD CDMRP Funding no
Abstract Background Despite improved outcomes with programmed cell death protein 1/programmed death-ligand 1 (PD-1/PD-L1) immune checkpoint inhibitors (ICI) in renal cell carcinoma (RCC) and non-small cell lung cancer (NSCLC), most patients develop resistance. Preclinical data demonstrate that Janus kinase inhibition (JAKi) restores T cell cytokine production and proliferation, potentiating ICI activity, with ruxolitinib plus ICI demonstrating superior tumor control across multiple murine models. Ruxolitinib may also reprogram immunosuppressive myeloid populations within the tumor microenvironment. In an investigator-initiated trial (NCT03681561), 53% of patients with relapsed/refractory Hodgkin’s Lymphoma previously failing anti-PD-1 therapy responded to ruxolitinib plus nivolumab, including 6 complete responses. High response rates were also observed with JAKi plus anti-PD-1 in NSCLC (NCT03425006). We aim to investigate whether combining ruxolitinib with anti-PD-1 therapy can overcome resistance in solid tumors. Methods This prospective, open-label, single-arm phase 1b study evaluates ruxolitinib plus retifanlimab in patients with advanced clear cell RCC or NSCLC progressing on prior PD-1/PD-L1 therapy. Patients must have measurable disease (Response Evaluation Criteria in Solid Tumors [RECIST] 1.1), Eastern Cooperative Oncology Group (ECOG) performance status 0-1, and adequate organ function. Key exclusions: prior JAKi, >1 prior line of PD-1/PD-L1 therapy, primary progression on prior ICI, or unresolved immune-related adverse event (>grade 1). A 3 + 3 dose escalation design evaluates ruxolitinib 15 mg and 20 mg twice daily with fixed retifanlimab 500mg intravenously (IV) every 4 weeks, followed by dose expansion. Treatment continues until progression, toxicity, or withdrawal. The primary objective is determining the recommended phase 2 dose (RP2D). Secondary objectives include objective response rate, duration of response, radiographic progression-free survival, overall survival, and safety. To further interrogate mechanisms of response and resistance, correlative studies evaluating tumor microenvironment, immune cell composition, and molecular profiling are ongoing. A Simon’s two-stage design is utilized for testing the null hypothesis (true response rate 10%) against an alternative of 30%, rejecting null if ≥ 6 responses occur in 25 participants (α = 0.05, power 80%). Assuming a dropout rate of 5%, we will target a minimum enrollment of 36 and maximum enrollment of 42 participants to ensure a maximum of 40 evaluable participants on the trial. The standard survival analysis will be conducted for the time-to-event endpoints. The study is activated at the University of California San Diego and is currently open to accrual. Clinical Trial Information: NCT07219576. Results Resistance to PD-1/PD-L1 blockade is a major unmet need in RCC and NSCLC, with few effective options after progression on frontline immunotherapy. PRISM tests a mechanistically rational combination, ruxolitinib to reverse chronic interferon-driven T cell dysfunction and immunosuppressive myeloid skewing, paired with retifanlimab to sustain antitumor T cell activity. Integrated correlative studies aim to identify biomarkers of response to guide patient selection. Our vision is to establish JAK inhibition as a tractable strategy for restoring checkpoint sensitivity across solid tumors, providing the foundation for a biomarker-driven randomized phase 2 trial in checkpoint-refractory disease. Conclusions NA
Abstract Background Therapeutic resistance to immune checkpoint inhibitors (ICIs) and antiangiogenic therapies remains a significant clinical challenge in advanced renal cell carcinoma (RCC). The prognostic value of CD8+ T-cell infiltration—a central player in cancer immunity—is often equivocal across RCC studies, supporting the need for a systematic characterization of response-associated T cell states. Here, we used a machine learning-based approach that integrated bulk RNA-sequencing (RNA-seq) data from multiple ICI-based clinical trials with an in-house single-cell RNA-sequencing (scRNA-seq) dataset to identify CD8+ T-cell states associated with progression-free survival (PFS), providing novel prognostic insights for advanced RCC. Methods We collected bulk RNA-seq data of baseline tumor samples and patient outcomes from four major clinical trials of advanced RCC with at least one treatment arm receiving ICI monotherapy or ICI combined with tyrosine kinase inhibitors (ICI+TKI): HCRN GU16-260, JAVELIN Renal 101, COSMIC-313, and CheckMate 9ER. An in-house scRNA-seq dataset of tumor-infiltrating CD8+ T cells from 70 advanced RCC patients was used as a cellular reference. Our analysis focused on the CD8+ exhausted T cells (Tex), which generally overlap with tumor-specific T-cell population. Using non-negative matrix factorization (NMF), we classified the Tex into four functional states. To link the scRNA-seq-derived T-cell states with the PFS and bulk RNA-seq data, we applied Scissor (Sun et al., 2022), a machine learning model that identifies phenotype-associated cell subsets through cell-sample correlations. For the bulk cohort meta-analysis, shared genes were normalized, scaled, and batch-corrected using ComBat (Johnson et al., 2007). Statistical significance was established via a permutation test of the PFS data, and calculating empirical p-values based on the concordance index (C-index). Results Our final bulk dataset comprised 1,527 RNA-seq tumor samples from four clinical trials of RCC, including 1,004 tumor samples from patients receiving ICI or ICI+TKI regimens. We first applied Scissor to each treatment arm of each trial individually, and found that worse PFS-associated Tex cells exhibited elevated expression of tissue-resident markers, such as ZNF683 and ITGAE, consistently across all ICI and ICI+TKI arms. These markers represent one specific NMF program defined in Tex, which we have termed the resident-memory-like exhausted state (Tex-rm). Notably, the Tex-rm population was enriched among worse PFS-associated cells in all ICI-based treatment arms (ratio of observed to expected [Ro/e] = 1.37-2.33) but not in any TKI monotherapy arms. Subsequently, we combined all ICI-based treatment arms and applied Scissor to the batch-corrected bulk dataset. This again identified Tex-rm as associated with worse PFS. Such a correlation was supported by a permutation test with p-value < 0.001. Conclusions Through machine learning-based integration of bulk and single-cell RNA-seq datasets, we found that baseline tumor infiltration by exhausted T cells with a resident memory phenotype (Tex-rm) is significantly associated with worse PFS in advanced RCC patients receiving ICI monotherapy or ICI+TKI combination therapy. This finding highlights the Tex-rm state as a potential predictive biomarker for ICI efficacy. DOD CDMRP Funding no
Abstract Background Local recurrence (LR) after nephrectomy for localized renal cell carcinoma (RCC) is rare but associated with poor oncologic outcomes. Definitions of LR remain inconsistent across studies, and atypical recurrences involving non-traditional intra-abdominal and incisional sites are poorly characterized. Emerging evidence suggests that minimally invasive surgical techniques may influence LR patterns. Here, we evaluate predictors and patterns of LR, with a specific focus on atypical presentations. Methods Using a prospectively maintained institutional kidney surgery database, we identified patients with localized RCC who underwent nephrectomy between 2000 and 2024 and had at least one post-operative cross-sectional imaging study. LR was defined as a new lesion on follow-up imaging and was classified according to criteria developed by Lee et al. This system defines the following: Type I, a single recurrence in the remnant kidney or ipsilateral renal fossa; Type II, a single recurrence involving the ipsilateral vasculature, ipsilateral adrenal, or regional lymph nodes; Type II, a single recurrence in other intra-abdominal soft tissues or organs; and Type IV, any combination of Types I-III or multiple of a single type. Predictors of LR-free survival were evaluated using multivariable Fine-Gray competing risks regression, with metastasis or death as competing events. We assessed predictors of any atypical (Type III or Type IV including Type III) recurrence using univariable Fine-Gray models. Secondary interaction analyses were performed to test whether the association between predictors and any atypical recurrence differed by surgical approach (open vs minimally invasive). Results Among 3,815 patients with localized RCC who underwent nephrectomy, 122 developed a LR, including 37 patients with an atypical (Type III) recurrence. At 20 years, cumulative incidence was 8.6% for any LR and 16% for metastasis; any Type III recurrence was rare at 2.5%. Positive surgical margins specifically in minimally invasive surgery were strongly associated with increased risk of LR (SHR 6.05; 95% CI 2.84, 12.9) and any Type III recurrence (SHR 9.03; 95% CI 3.20, 25.5). Other predictors included higher tumor stage, high grade, and partial nephrectomy. Conclusions LR after nephrectomy was uncommon, including atypical presentations, which rarely occurred before other competing events. Overall, predictors of LR were similar for minimally invasive and open approaches; however, positive surgical margins were strongly associated with overall and atypical LR specifically after minimally invasive surgery. Higher stage and grade were independently associated with increased LR risk. These findings highlight the predominant role of tumor biology in determining recurrence risk. Furthermore, surgeons should make every effort to avoid positive surgical margins during nephrectomy by any technique.
Abstract Background Chromophobe Renal Cell Carcinoma (ChRCC), derived from the intercalated cells of the collecting duct, typically follows an indolent clinical course. However, approximately 5–8% of cases undergo sarcomatoid transformation (sChRCC), a high-grade dedifferentiation resulting in a spindle-cell morphology and a rapid, metastatic trajectory. This transition represents a profound clinical shift; while classic ChRCC (cChRCC) is often surgical, sChRCC is notoriously refractory to anti-angiogenic agents and tyrosine kinase inhibitors (TKIs). The molecular drivers and spatial architecture of this transition have remained elusive due to the rarity of the subtype and the limitations of bulk genomic approaches that mask the spatial crosstalk between malignant cells and the surrounding stroma. Methods We established a high-resolution spatial transcriptomic atlas of ChRCC using the Xenium 5K platform, profiling n = 6classic and n = 3 sarcomatoid tumor samples. We integrated this subcellular spatial data with bulk RNA-seq and validated findings through immunohistochemistry (IHC) for markers such as CD8, CD68, and CD163. Using reference-based mapping and intercellular communication analysis, we dissected the “niche-specific” interactions that define the sarcomatoid tumor microenvironment (TME). Results Our spatial atlas revealed a profound cellular reorganization during the sarcomatoid transition. Compositional analysis showed that while cChRCC is predominantly “immune-excluded,” sChRCC displays a massive enrichment of T cells and macrophages deeply infiltrated within mesenchymal bundles. We identified that sChRCC tumor cells undergo coordinated mesenchymal reprogramming, driven by a robust Epithelial-Mesenchymal Transition (EMT) signature (e.g., ADAM12, POSTN, COL5A1) and a distinct dysregulation of ferroptosis-related metabolic pathways. We observed a paradoxical “infiltration-exhaustion” state: despite high T-cell density, these cells exist in an advanced state of exhaustion, expressing high levels of CTLA4, TIGIT, LAG3, and HAVCR2. Spatial mapping confirmed the co-localization of these receptors with the inhibitory ligand CD276 (B7-H3) on sarcomatoid tumor cells and cancer-associated fibroblasts (CAFs), suggesting a dominant spatial mechanism of immune evasion. Intercellular communication analysis nominated the CXCL12-CXCR4 axis as a primary “spatial sculptor.” While cChRCC signaling is sparse and focused on vascular integrity (VEGFA-VEGFR2), sChRCC is characterized by a dense, complex network where tumor cells and CAFs serve as central signaling hubs. We identified a unique CXCL12 gradient derived from tumor cells that acts as a chemoattractant for T cells and M2-polarized macrophages (CD163+, MRC1+). Furthermore, a novel plasma cell-to-tumor signaling hub via the GAS6-AXL axis was identified as a driver of EMT progression. Physical remodeling was evidenced by a massive increase in collagen-integrin signaling (e.g., COL1A1-CD44), which effectively “traps” the immune infiltrate within a pro-tumorigenic fibrotic niche. Conclusions This study provides the first definitive structural blueprint of the sChRCC microenvironment. We demonstrate that the sarcomatoid transition is not merely a cell-intrinsic morphological shift, but a wholesale reorganization of the tumor ecosystem into a highly coordinated, immunosuppressive niche. By mapping these “Achilles’ heels,” we move beyond simple genomic mutations to nominate B7-H3, the CXCL12-CXCR4 axis, and AXL signaling as spatially localized therapeutic vulnerabilities. These findings suggest that future clinical efforts for sChRCC must pivot from traditional anti-angiogenic strategies toward multi-modal immunotherapies that simultaneously disrupt mesenchymal reprogramming, inhibit suppressive myeloid recruitment, and reinvigorate the sequestered, exhausted T-cell population. This spatial framework offers a biologically informed path toward personalized treatment for this lethal ChRCC variant.
