About

Tyler Stewart is an Associate Clinical Professor in the Department of Medicine at UC San Diego. His research focuses on urothelial carcinoma, prostate cancer, and bladder cancer, with an emphasis on the manipulation of the gut microbiome, immunotherapy, and the molecular and genetic aspects of cancer treatment. He has contributed to studies on the impact of non-canonical WNT signaling, tumor mutational burden, microsatellite instability, and circulating tumor DNA in urothelial carcinoma. His work also includes investigating the use of immune checkpoint inhibitors, oncolytic adenoviral therapy, and the role of germline testing in at-risk cancer patients. Stewart's research aims to improve treatment outcomes and develop innovative therapeutic strategies for genitourinary cancers.

Research topics

• Oncology
• Medicine
• Internal medicine

Selected publications

• Induction enfortumab vedotin plus pembrolizumab followed by maintenance pembrolizumab in first-line metastatic urothelial carcinoma (IMPROEV).

  Journal of Clinical Oncology · 2026-03-01

  article

  TPS893 Background: Enfortumab vedotin plus pembrolizumab (EV/P) is the preferred first-line regimen for metastatic urothelial carcinoma (mUC), demonstrating significant improvements in overall survival (OS) and progression-free survival (PFS) compared with platinum-based chemotherapy. However, continuation until disease progression or unacceptable toxicity leads to cumulative adverse events, particularly peripheral neuropathy, which is a common cause of treatment discontinuation. Post-hoc exposure–response analyses indicated that early EV dose intensity was associated with response, while declining serum EV concentrations over time did not compromise efficacy. In EV-103 and EV-201, 38–50% of patients who discontinued EV or EV/P maintained disease control for years, suggesting durable benefit after treatment cessation. These data support a finite-duration, induction–maintenance strategy over the traditional “treat-to-toxicity” approach. We hypothesize that induction EV/P followed by pembrolizumab maintenance in responders will preserve durable oncologic control while reducing cumulative toxicity and improving quality of life. Methods: IMPROEV is a single-arm, open-label, nonrandomized phase II trial enrolling 97 previously untreated patients with locally advanced or metastatic UC across four centers. Patients who received prior neoadjuvant or adjuvant checkpoint inhibitor therapy ≥12 months before enrollment are eligible. Patients receive EV 1.25 mg/kg IV on days 1 and 8 plus pembrolizumab 200 mg IV on day 1 of each 21-day cycle for 6 cycles (18 weeks). Patients achieving confirmed complete or partial response per RECIST v1.1 transition to P 400 mg IV every 6 weeks for up to 2 years. Those with stable disease may continue EV/P or transition to P alone at investigator discretion. The primary endpoint is 18-month PFS. Secondary endpoints include OS, duration of response, treatment-free interval, incidence of peripheral neuropathy, and quality of life (EORTC QLQ-C30, BLM30, CIPN20). Exploratory analyses will evaluate ctDNA dynamics from baseline through progression. The trials uses a one-arm, two-stage design testing the null hypothesis that the proportion of patients who progress or die within 18 months ( p 0 ) is ≤0.56 versus the alternative ( p a ) of 0.67, with 11.4% type I error and 80.3% power. Stage I will enroll 31 evaluable patients; if >22 progress or die within 18 months, accrual will stop. Otherwise, 66 additional patients will be enrolled in stage II. Enrollment is expected to begin in Fall 2025. Clinical trial information: NCT07221942 .

  PublisherDOI

• Validation of a computational histology artificial intelligence (CHAI) prognostic biomarker in muscle invasive bladder cancer.

  Journal of Clinical Oncology · 2026-03-01

  article

  791 Background: Standard management for MIBC is radical cystectomy (RC). This traditionally includes neoadjuvant chemotherapy (NAC) which confers additional survival benefit. However, the absolute survival benefit of NAC is 5-10%, and only 30-40% of US MIBC cases receive NAC. These limitations underscore the need for a more personalized approach to identify patients most likely to benefit from treatment escalation or avoid overtreatment in those likely cured with RC alone. The CHAI biomarker (BM) platform has previously been used to develop a now commercially-available prognostic tool in non-MIBC. We aimed to validate a novel prognostic BM, previously developed specifically for MIBC using the CHAI platform, in a multi-institutional real-world (RWD) MIBC cohort. Methods: The CHAI platform applies deep-learning to extract quantitative histologic features from pre-treatment transurethral resection bladder tumor (TURBT) specimen H&E-stained whole-slide images. The previously-developed MIBC BM was locked; the model outputs a continuous histologic risk signature and is dichotomized into unfavorable BM positive (BM+) and favorable BM negative (BM-) groups. Validation was done on an independent, held-out, retrospective, pooled, RWD of clinical T2N0M0 MIBC patients who underwent RC at multiple NCI-Designated Centers. Multivariate (MVA) Cox proportional hazards (CPH) models assessed associations with recurrence-free survival (RFS), cause-specific survival (CSS), and OS. Kaplan-Meier methods and log-rank tests were used for survival analysis. Results: Among 134 patients with MIBC, 67 (50%) patients received cisplatin-based NAC. The BM stratified patients into 67 (50%) BM+ and 67 (50%) BM-. Among patients who received NAC, pathologic complete response (pCR) was seen at RC among 6/29 (20%) BM+ vs 13/38 (34%) BM-. On MVA, controlling for age, sex, presence of carcinoma in situ (CIS), variant histology, and NAC, BM+ was significantly associated with inferior RFS (Table; HR: 2.70 [1.46, 4.97]), CSS (HR: 2.60 [1.29, 5.26]) and OS (2.29 [1.29, 4.04]), all p<0.01. With 36-mo median follow up, at 2yr, BM+ had worse outcomes: RFS 50% vs 18%; CSS 25% vs 8.3%; OS 36% vs 15%. Conclusions: A histologic prognostic BM derived from pre-treatment H&E TURBT slides was validated, stratifying clinical T2 MIBC patients by risk of RFS, CSS, and OS, even when controlling for NAC. Future efforts will evaluate the ability to predict which patients benefit the most from additional treatment. Such a tool could be used to optimize patient selection by identifying those most likely to benefit from perioperative therapies, and sparing those who may be cured with surgery alone. MVA for RFS. HR (95% CI) P value Biomarker 2.70 [1.46, 4.97] p=0.01* Age 1.03 [1.00, 1.07] p=0.06 Sex 1.24 [0.65, 2.35] p=0.52 CIS 0.69 [0.32, 1.50] p=0.35 Variant histology 1.15 [0.56, 2.35] p=0.71 NAC 1.54 [0.80, 2.95] p=0.20

  PublisherDOI

• Anxiety, Depression, and Suicidality Among Testicular Cancer Survivors

  Cancer Medicine · 2026-02-01

  articleOpen access

  INTRODUCTION: We evaluated the incidence of anxiety, depression, and suicidality amongst TC survivors and the impact of chemotherapy on these outcomes. METHODS: We conducted a retrospective cohort study of men diagnosed with TC in the United States Veterans Affairs Health System from 1990 to 2016. De novo anxiety or depression was a composite endpoint comprised of diagnosis codes for anxiety, depression, or medications used to treat these diagnoses. Incident suicidality was defined as a diagnosis code for suicidal ideation. 2022 TC patients were compared in a 1:3 ratio to 6375 controls. Cox proportional hazards models were employed for statistical analysis. RESULTS: Mean age at diagnosis was 42.6 years. 5-year cumulative incidence of anxiety or depression was 53.4% in TC patients and 35% for controls (p < 0.001). TC patients were more likely to develop anxiety or depression (HR 1.66, 95% CI 1.56-1.78, p < 0.001) and suicidality (HR 22.99, 95% CI 17.52-30.17, p < 0.001). In the TC cohort, factors associated with a higher risk of anxiety or depression were divorce (HR 1.15, 95% CI 1.00-1.32, p = 0.044), unemployment (HR 1.68, 95% CI 1.47-1.9, p < 0.001), and receipt of chemotherapy (HR 1.20, 95% CI 1.06-1.35, p < 0.001). CONCLUSIONS: Psychological morbidity due to depression, anxiety, and suicidality is high among TC survivors. In our analysis chemotherapy increases the rates of psychosocial morbidity. Clinicians should be proactive in screening and intervening for these diagnoses in TC survivors to provide early intervention and improve health comes.

  PublisherDOI

• Randomized pilot trial of pre-visit patient education as preparation for decision making in newly diagnosed prostate cancer (PrepDM).

  Journal of Clinical Oncology · 2026-03-01

  article

  361 Background: Men with newly diagnosed prostate cancer (PCa) face preference-sensitive choices among currently available alternatives in management. High-quality decisions require understanding options, weighing trade-offs, and aligning choices with personal values. We evaluated whether brief pre-visit education improves patients’ Preparation for Decision Making (PrepDM). Methods: This was a single-center, 1:1 randomized pilot trial of men with biopsy-proven prostate cancer allocated to intervention (in which patients were provided curated educational material/videos before the follow-up visit) or usual care. Baseline demographics and tumor characteristics were collected; continuous and categorical variables were compared with Wilcoxon rank sum and Fisher’exact /Chi-square tests, respectively. The primary outcome was the PrepDM total score (10 items, each 1-5; range 1-50), assessing overall patient degree of strategy understanding and preparation. Secondary outcomes were item-level scores. Analyses were intention-to-treat. Group differences used Wilcoxon rank-sum with Hodges-Lehmann (HL) estimates; item-level multiple comparison testing was controlled with Benjamini-Hochberg FDR (q-values). Results: Fifty-two participants were randomized (26/arm). Median age was 70 years; median PSA 7.2 ng/dL; median prostate volume 42 cc; ISUP GG1/GG2/≥GG3 were 27%/37%/36%. Cohorts were comparable at baseline with no detectable differences in race/ethnicity, PSA, Gleason grade group, prostate volume, BMI, or age. The intervention increased overall preparation compared to control (total score difference: +6 points, 95%CI +2 to +10; p=0.002). After FDR adjustment, 7/10 items favored intervention: helping organizing thoughts (+1; q=0.006); knowing the decision depends on what matters most (+1; q=0.02); deciding desired involvement (+1; q=0.02); identifying questions (+1; q=0.03); preparing to talk about what matters most (+1; q=0.04); recognizing a decision is needed (+1;q=0.041); and weighing pros/cons (+1; q=0.045). Items on preparing for a follow-up visit (q=0.08), making a better decision (q=0.11), and prioritizing which pros/cons are most important (q=0.11) were not significant. Conclusions: Brief pre-visit patient education improved preparation for decision making, supporting shared decision-making and patient empowerment in prostate cancer care. Larger multicenter trials should confirm effectiveness and assess downstream outcomes (decisional conflict, decision quality, and satisfaction/regret).

  PublisherDOI

• BPI26-018: Integrated Specialty Prostate Cancer Care: Outcomes From a Collaborative Clinic Model at UC San Diego Moores Cancer Center

  Journal of the National Comprehensive Cancer Network · 2026-03-27

  article
  PublisherDOI

• Beyond Sterility: the Urinary Microbiome in Bladder Cancer Carcinogenesis and Treatment

  Current Urology Reports · 2026-01-14

  article
  PublisherDOI

• Prognosis of patients (pts) with synchronous vs metachronous metastatic or locally advanced urothelial carcinoma (la/mUC).

  Journal of Clinical Oncology · 2026-03-01

  article

  695 Background: There is relatively limited data regarding the prognostic role of the presence of synchronous metastasis or de novo locally advanced (unresectable) tumor in pts with UC. Ιn this retrospective study, we examined the prognosis of pts with synchronous vs metachronous la/mUC (prior localized tumor) in a real-world cohort. We hypothesized that pts with synchronous la/mUC would have worse prognosis. Methods: We included pts from 26 centers in US and Europe treated with anti-PD(L)1 immune checkpoint inhibitors (ICI) in any therapy setting. We calculated overall survival (OS) from the time of initial diagnosis for pts with synchronous la/mUC, and from the time of diagnosis of advanced disease for pts with metachronous la/mUC, until death using Kaplan-Meier method. Pts with diagnosis of la/mUC ≤12 weeks from the initial UC diagnosis and without receiving any interim treatment, were assigned to the synchronous la/mUC group. We also calculated OS, progression-free survival (PFS) from 1st line (1L) Immune Checkpoint Inhibitor (ICI) start until progression or death for PFS, and until death for OS, as well as overall response rate (ORR) in pts with synchronous vs metachronous la/mUC receiving 1L ICI treatment. Multivariable (MVA) models were adjusted using the Khaki risk factors (ECOG PS ≥2, albumin &lt; 3.5 g/dL, neutrophil/lymphocyte ratio &gt; 5, presence of liver metastases). Results: We identified 1537 pts with la/mUC with median age of UC diagnosis 69 years (75% men, 81% White, 20% upper tract primary tumor, 71% pure UC, 40% visceral metastasis (liver/lungs), 75% ECOG PS 0-1, prior platinum-based chemotherapy [19% as neoadjuvant, 46% as 1L metastatic]). Median follow up from time of initial UC diagnosis was 42 months (mo); 511 pts had synchronous la/mUC and 1026 had prior localized tumor. Median ΟS was 22 (95%CI 20-25) mo with synchronous la/mUC vs 25 (95%CI 23-28) mo with metachronous la/mUC (HR 1.18, 95%CI 1.0-1.4, p = 0.045). OS, PFS and ORR MVA analyses with 1st line ICI monotherapy are shown in Table. Conclusions: We found that synchronous la/mUC was associated with worse prognosis vs metachronous la/mUC, but further validation is needed. Limitations include retrospective design, lack of randomization, selection and confounding biases, while 1L therapy did not include enfortumab/pembrolizumab combination. Our hypothesis-generating data could inform prognostic estimates and stratification factors for clinical trials. N mOS, mo(95%CI) HR([95%CI], p) N PFS, mo(95% CI) HR([95%CI], p) N ORR %([95%Cl]) OR(95%CI, p) 1L ICI Synchronous la/mUC 200 15 (10-20) 1.06 ([0.8-1.4], 0.64) 154 4 (2-6) 0.87 ([0.67-1.1], 0.29) 189 32% (25-38) 1.2 ([0.8-1.08], 0.39) Metachronous la/mUC 560 18 (15-21) ref 439 4 (3-5) ref 541 30% (26-34) ref

  PublisherDOI

• Killing cancer takes guts: lessons learned from the manipulation of gut microbiome and immunotherapy for the future of urothelial carcinoma

  OncoImmunology · 2026-01-21 · 1 citations

  articleOpen accessSenior author

  Urothelial carcinoma (UC) remains a common cancer with significant morbidity and mortality worldwide. Immune checkpoint inhibitors (ICIs) have helped revolutionize the treatment of UC, and there is growing evidence suggesting the crucial role of the gut microbiome in immune system function influences immunotherapy outcomes in this disease. Herein, we review the preclinical basis for how manipulation of the gut microbiome may alter the efficacy of immunotherapy for patients with cancer, highlight interventions optimizing gut microbiome diversity currently in use, review recent and ongoing clinical trials supporting the role of the gut microbiome in improving immunotherapy outcomes, and discuss clinical implications to improve outcomes for UC patients with immunotherapy in the real world. There is growing evidence that suggests that specific gut microbiome compositions significantly modulate the host immune system and response to ICIs. Early studies have shown that certain microbial taxa enhance antitumor immunity by influencing T cell priming, dendritic cell activation, and cytokine production. Fecal microbiota transplantation (FMT), probiotic supplementation, and dietary modulation have emerged as promising methods to alter microbiomes to improve immunotherapy outcomes. Taxa from positive immunotherapy responders across a variety of cancers demonstrate beneficial effects when transplanted into both treatment-naive or prior nonresponders. Increasing evidence suggests that the gut microbiome plays a crucial role in cancer care, particularly when patients are treated with immunotherapy. Future studies are needed to better understand the underlying mechanisms. While some studies are currently underway to explore gut manipulation for patients with UC, more studies are needed to investigate the potential to convert nonresponders into responders through microbiome manipulation.

  PublisherDOI

• Real-world analysis of 2IR immune response score in urothelial carcinoma (UC).

  Journal of Clinical Oncology · 2025-02-10 · 1 citations

  article

  739 Background: Immune checkpoint inhibition (ICI) is often used to treat UC, however many patients (pts) exhibit resistance. Using bulk RNA sequencing of pre-treatment tissue from ImVigor 210 and CheckMate 275 trials, an immune response score (2IR) was previously developed to predict ICI response by determining the comparative expression of genes involved in anti-tumorigenic adaptive immune response and pro-tumorigenic inflammation. Here, we evaluated the 2IR score as a prognostic or predictive biomarker in real-world pts with UC. Methods: Specimens from pts with UC (n = 6395) were profiled at Caris Life Sciences (Phoenix, AZ) using next generation sequencing (NGS) of DNA and RNA. 2IR was calculated by comparative RNA expression of 10 adaptive immune genes and 39 pro-tumorigenic genes. Tumors were classified as 2IR-Low (2IR ≤ -0.5), -Mid (-0.5 &lt; 2IR &lt; 0), and -High (2IR ≥ 0) as previously described. Spearman correlation analysis was utilized to compare 2IR with PD-L1 combined proportion score (CPS), tumor mutation burden (TMB), interferon score (IFN) and the tumor microenvironment (TME) cell fractions estimated using quanTIseq. Clinical outcomes included real-world overall survival (OS) from ICI start to last contact and time on treatment (ToT) with pembrolizumab, obtained via matched insurance claims data and calculated using Kaplan-Meier methods while Hazard ratio (HR) was calculated by Cox proportional model. Results: Of the 6395 UC samples, 9.5% (n = 611) were classified as 2IR-High, 42.7% (n=2730) 2IR-Mid, and 47.8% (n=3054) 2IR-Low. In pts with Upper Tract UC primary (UTUC, n = 1104), 2IR was comparatively lower than in pts with bladder cancer primary (BC, n = 4923) (median: -0.53 vs. -0.47, p &lt; 0.001). 2IR shows a positive correlation with TMB (r = 0.23; p &lt; 0.001), IFN score (r = 0.21; p &lt; 0.001) and insignificant correlation with PD-L1 CPS (r = 0.09; p &lt;0.001). 2IR-High positively correlated with mutations in TP53 , RB1 , and ARID1A (all p &lt; 0.0001), amplification of ERBB2 (p &lt; 0.0001), and microsatellite-instability high status (p &lt; 0.0001), but not with FGFR3 mutations or fusions. 2IR positively correlated with immune cell fractions, such as CD4 + T (r = 0.09), CD8 + T (r = 0.22), and dendritic cell (r = 0.18), but negatively correlated with B cells (r = -0.06), M1 macrophages (r = -0.25), M2 macrophages (r = -0.08) and neutrophils (r = -0.12). Pts treated with 2IR-High tumors had longer ToT on pembrolizumab (HR = 0.65 CI 0.55 - 0.77, p &lt;0.001) and OS from ICI start (HR = 0.51, CI 0.41 – 0.62, p &lt; 0.001) compared to pts with 2IR-Low UC. This association held for both primary BC and UTUC. Conclusions: In a real-world cohort, we validate the 2IR score and suggest that it is prognostic for OS and predictive for pembrolizumab ToT, consistent with a positive correlation with known predictors of ICI response. Prospective studies are needed to further validate this biomarker for use in clinical practice, especially given the evolving UC treatment landscape.

  PublisherDOI

• Natural history and risk-stratification of biochemical recurrence in prostate cancer treated with definitive radiotherapy.

  Journal of Clinical Oncology · 2025-02-10

  article

  348 Background: Biochemical recurrence (BCR) of prostate cancer following radiation therapy often does not result in clinically significant events, but a subset of patients may be at higher risk of developing metastatic disease or death. Identifying these patients is critical for clinical decision making, but unlike in the post-prostatectomy setting, no method has been externally validated for BCR after radiation therapy. This study characterized outcomes after post-radiation BCR and validated a proposed risk stratification heuristic. Methods: This was a retrospective, multicenter, nationwide cohort study of patients having post-radiation BCR treated in the United States Veterans Administration Health System. High-risk post-radiation BCR was defined as either Gleason score ≥8 or BCR occurring within 18 months of radiation therapy. BCR was defined as post-treatment PSA greater or equal to PSA nadir + 2 ng/ml, initiation of androgen deprivation therapy distinct from the initial treatment course, or development of metastatic disease, whichever occurred first. Results: Median time to BCR was 42.5 months (interquartile range 22.9-73.0). Among 7,126 patients who experienced BCR, 35.5% of patients developed metastatic disease and 17.4% died of prostate cancer at 10 years. 38.5% of patients had at least one qualifying high-risk feature of whom 23.3% had high-risk recurrence based on time to recurrence alone, 57.6% based on Gleason/Grade Group alone, and 19.1% based on both criteria. High-risk BCR resulted in higher 5-year rates of metastatic disease (42.0% versus 24.5%, hazard ratio (HR) 1.83, 95% confidence interval (CI) 1.69-1.98, p &lt; 0.001), death from prostate-cancer (18.7% versus 8.78%, HR 1.82, 95% CI 1.63-2.03, p &lt; 0.001), and death from all causes (37.1% versus 30.8% rates, HR 1.18, 95% CI 1.11-1.26, p &lt; 0.001). Conclusions: A simple, two-element risk stratification tool using existing clinically data is the first validated tool for identifying patients at risk of metastases or PCSM following post-radiation BCR. Most patients experiencing BCR in this context do not develop metastases or lethal prostate cancer, making such stratification essential for treatment decision-making and refinement of patient populations for clinical trials. Further work remains on risk-adapted therapy intensification.

  PublisherDOI

Frequent coauthors

• Arjun Gupta

  106 shared

• Chad D. Rethorst

  Center for Advancing Health

  105 shared

• Muhammad S. Beg

  The University of Texas Southwestern Medical Center

  104 shared

• Kimberli Crane

  The University of Texas Southwestern Medical Center

  101 shared

• Zain Rahimi

  Merrimack Pharmaceuticals (United States)

  101 shared

• Ying Dong

  Shenzhen University

  100 shared

• Nizar Bhulani

  100 shared

• Brent S. Rose

  University of California, San Diego

  66 shared

Education

• Ph.D., Molecular and Computational Biology

  University of California, San Diego

  2005

• M.S., Molecular and Computational Biology

  University of California, San Diego

  2001

• B.S., Biology

  University of California, San Diego

  1999