Supplemental Table S2. Immunohistochemistry Results and Clinical Benefit

There has been a controversy about the predictive value of tissue-TMB-H for immune checkpoint inhibitors (ICIs) with limited data regarding blood-TMB (bTMB) in GI tumors. We aim to evaluate the predictive value of bTMB compared with MSI-H in GI tumors. Patients with unresectable/metastatic GI cancer, harboring either MSS with bTMB-H (≥10 mut/Mb) or dMMR/MSI-H who received ICI were included. We compared ICIs’ efficacy between MSS-bTMB-H (N=45) versus MSI-H (N=50) in GI tumors. Ninety-five patients were identified with the majority having colorectal (49.5%) or esophagogastric (34.7%) cancers. MSS-bTMB-H group had more esophagogastric cancer and later-line ICI recipients, with no significant differences in other known prognostic variables. At a median follow-up of 9.4 months, MSI-H group showed superior ORR (58.0% vs. 26.7%), DCR (84.0% vs. 42.2%), DoR (not-reached vs. 7.6 mo), PFS (22.5 vs. 3.8 mo), and OS (Not-reached vs. 10.1 mo) compared with MSS-bTMB-H. Multivariable analysis showed that MSI-H was an independent favorable factor over MSS-bTMB-H for PFS (HR=0.31, CI 0.15-0.63, P =0.001) and OS (HR=0.33, CI 0.14-0.80, P =0.014). MSI-H group showed favorable outcomes compared with MSS-bTMB-16+ (ORR: 58.0% vs. 26.9%; DCR: 84.0% vs. 42.3%; PFS:22.5 vs. 4.0 mo) and MSS-bTMB-20+ (ORR: 58.0% vs. 31.6%; DCR: 84.0% vs. 42.1%; PFS:22.5 vs. 3.2 mo). There was no difference between MSS-bTMB10–15 compared with MSS-bTMB-16+ in ORR, DCR, and PFS, or between MSS-bTMB10–19 compared with MSS-bTMB20+. Regardless of bTMB cutoff at 10, 16, or 20, bTMB-H did not appear to be a predictive biomarker in MSS GI tumors in this retrospective analysis.

Abstract Background Papillary renal cell carcinoma (pRCC) accounts for up to 15% of all kidney cancer cases, yet our understanding of its tumor immune microenvironment (TIME) remains limited. We utilized multiplex immunofluorescence (mIF) and spatial transcriptomics (ST) to evaluate immune cell spatial architecture in pRCC and compared to that of clear cell RCC (ccRCC). Methods Surgical tumor specimens were obtained from localized RCC tumors, followed by mIF using markers for T cells, B cells, and tumor-associated macrophages (TAMs). Spatial data were derived in regions of interest (ROIs) manually selected from spatially distinct tissue compartments of the TIME. Single-cell ST was performed on a subset of patient samples, utilizing probes against 960 transcripts. Cell abundance, cell spatial clustering, and spatially varying gene expression were analyzed to identify unique features of the TIME in pRCC. Results Sixteen pRCC and 70 ccRCC patient samples underwent mIF. Compared to ccRCC, global pRCC immune cell abundance was statistically lower amongst functional CD8 T cells, while global cell spatial clustering was higher amongst M2-like macrophages as measured by mIF, including PDL1+ subsets. Using ST, seven genes were significantly associated with spatial clustering of M2-like macrophages in pRCC. Three of seven genes (CCL18, GPNMB, CD9) are known markers of lipid-associated TAMs (LAMs) (adjusted p-value < 0.1). Conclusions Compared to ccRCC, pRCC has fewer T cells but greater M2-like macrophage spatial clustering. Using ST, we found that multiple LAM-associated genes are spatially enriched in pRCC. Additional resources should be dedicated to investigating myeloid biomarkers and the impact of myeloid modulating therapeutics in pRCC.

Supplementary Table S1. Representativeness of Study Participants

AbstractPurpose: The combination of gemcitabine and docetaxel is often used to treat patients with recurrent osteosarcoma. Nab-paclitaxel has preclinical activity against osteosarcoma and is potentially less myelosuppressive than docetaxel. We conducted a prospective multi-institutional phase II trial combining gemcitabine and nab-paclitaxel for patients aged 12 to 30 years with recurrent osteosarcoma and measurable disease. Patients and Methods: A Simon’s two-stage design was used to test a 4-month progression-free survival (PFS-4) of 10% vs. 35%. Patients received nab-paclitaxel 125 mg/m2 and gemcitabine 1,000 mg/m2 weekly × 3 in 4-week cycles. Immunohistochemical analysis of archival tissue and serial assessment of circulating tumor cells (CTC) and circulating tumor DNA (ctDNA) using ultralow passage whole-genome sequencing were performed to identify potential biomarkers of response. Results: Eighteen patients received 56 total cycles (median 2, range 1–12). Two patients (11%) experienced confirmed partial response and six (33%) received >2 cycles. The PFS-4 was 28% (95% confidence interval, 13%–59%). Six patients required dose reductions and three patients were removed due to toxicities. All 18 patients had detectable CTCs and 10 had ctDNA identified. All eight patients with MYC amplification at study entry experienced disease progression. Conclusions: Gemcitabine and nab-paclitaxel demonstrated similar clinical activity and toxicity compared to previous retrospective reports utilizing gemcitabine and docetaxel in patients with recurrent osteosarcoma. Serial analysis of CTC and ctDNA was feasible in this prospective multi-institution study and provides preliminary data on the use of these assays in patients with relapsed disease.

A specific splicing isoform of RNASET2 is associated with worse oncologic outcomes in clear cell renal cell carcinoma (ccRCC). However, the interplay between wild-type RNASET2 and its splice variant and how this might contribute to the pathogenesis of ccRCC remains poorly understood. We sought to better understand the relationship of RNASET2 in the pathogenesis of ccRCC and the interplay with a pathogenic splicing isoform (RNASET2-SV) and the tumor immune microenvironment. Using data from The Cancer Genome Atlas and Clinical Proteomic Tumor Analysis Consortium, we correlated clinical variables to RNASET2 expression and the presence of a specific RNASET2-SV. Immunohistochemical staining with matched RNA sequencing of ccRCC patients was then utilized to understand the spatial relationships of RNASET2 with immune cells. Finally, in vitro studies were performed to demonstrate the oncogenic role of RNASET2 and highlight its potential mechanisms. RNASET2 gene expression is associated with higher grade tumors and worse overall survival in The Cancer Genome Atlas cohort. The presence of the RNASET2-SV was associated with increased expression of the wild-type RNASET2 protein and epigenetic modifications of the gene. Immunohistochemical staining revealed increased intracellular accumulation of RNASET2 in patients with increased RNA expression of RNASET2-SV. In vitro experiments reveal that this accumulation results in increased cell proliferation, potentially from altered metabolic pathways. RNASET2 exhibits a tumor-promoting role in the pathogenesis of ccRCC that is increased in the presence of a specific RNASET2-SV and associated with changes in the cellular localization of the protein.

Supplemental Table S3. Detailed ctDNA results over the course of the study.

BACKGROUND:Uveal melanoma often metastasizes to the liver, portending a poor prognosis. Melphalan/hepatic delivery system (HDS) via percutaneous hepatic perfusion (PHP) is a minimally invasive means of circulating high-dose chemotherapy through the affected liver. This study evaluated melphalan/HDS use as either first-line or second-line treatment to guide treatment sequencing. PATIENTS AND METHODS:A retrospective review included patients with hepatic-dominant metastatic uveal melanoma who underwent melphalan/HDS treatment via PHP from 2008 to 2023. RESULTS:A total of 30 patients were identified; 53.3% female, with a median age of 63.5 years (37-78 years). Median follow-up time was 14.5 months. First-line therapies included melphalan/HDS (n = 17), liver-directed (n = 7), and immunotherapy (n = 6). Second-line therapies included melphalan/HDS (n = 6), immunotherapy (n = 5), and liver-directed (n = 3). Median hepatic progression-free survival (hPFS) for first-line melphalan/HDS, immunotherapy, and liver-directed therapy was 17.6/8.8/9.2 months, respectively (P = 0.002). Median hPFS for second-line melphalan/HDS, immunotherapy, and liver-directed therapy was not reached/14.7/7.5 months, respectively (P < 0.001). Median overall PFS for first-line melphalan/HDS, immunotherapy, and liver-directed therapy was 15.4/8.8/9.2 months, respectively (P = 0.04). Median overall PFS for second-line melphalan/HDS, immunotherapy, and liver-directed therapy was 22.2/14.7/7.5 months, respectively (P = 0.001). CONCLUSIONS:Melphalan/HDS via PHP for metastatic uveal melanoma to the liver was found to have significantly improved hPFS and overall PFS when used as first-line therapy compared with immunotherapy or liver-directed therapy. PHP continued to demonstrate improved hPFS and PFS when used as second-line therapy compared with second-line immunotherapy or liver-directed therapy.

Background Both TMB-H and dMMR/MSI-H have been proposed as predictive biomarkers and received tissue-agnostic approval for CPIs, but there has been no head-to-head comparison between TMB-H and MSI-H as predictive biomarkers for CPIs. MSI-H demonstrated its promising predictive value in multiple independent cohorts with different CPI agents. In contrast, TMB-H has been suggested as a CPI predictive biomarker mainly based on the exploratory study from KEYNOTE-158, which showed response benefit without further meaningful clinical endpoints, raising concern for the validity of TMB-H as a predictive biomarker for CPIs. Also, data for TMB-H in GI cancer is limited. Methods We reviewed Moffitt Cancer Center databases for patients with histologically confirmed advanced/metastatic GI cancer, harboring either MSS with TMB-H (≥ 10 mut/Mb by tissue and/or blood) or dMMR/MSI-H (by IHC, PCR or NGS) who received CPI between June 2016 and April 2023 with measurable disease by RECIST1.1. We compared CPIs' efficacy profiles including ORR, DCR, DoR, PFS, and OS between MSS-TMB-H (N=56) versus MSI-H (N=50) in GI tumors. Results 106 patients with 50(47.2%) colorectal, 38(35.8%) esophagogastric, 11(10.4%) pancreaticobiliary and 7(6.6%) other GI tumors were included. MSS-TMB-H group had more later-line CPI recipients and esophagogastric cancer, but otherwise there was no significant difference in key prognostic variables including age, ECOG-PS, brain metastasis, and CPI type (table 1). At a median follow-up of 9.4 months, CPI-recipients with MSI-H GI tumors showed significantly superior ORR (58.0% vs 32.1%, p=0.0132), DCR (84.0% vs 50.0%, p=0.0005), PFS (22.5mo vs 4.8mo, p=0.0007), OS (Not reached vs 12.6mo, p=0.0004), and statistically non-significant trend towards improved DoR (Not reached vs 10.6mo, p=0.1669) compared to MSS-TMB-H (figures 1 and 2). Multivariable analysis with adjusting key prognostic variables including lines of treatment showed that MSI-H was an independent favorable factor compared to MSS-TMB-H for PFS (HR=0.50, CI 0.26–0.95, p=0.033) and OS (HR=0.34, CI 0.14–0.78, p=0.011). In the following subgroup analyses, MSS-TMB-20+ (≥20 mut/Mb) showed statistically non-significant trend towards improved ORR, DCR, and PFS compared to MSS-TMB-10s (20>TMB ≥10) (figures 1 and 3). Overall, MSI-H group showed favorable ORR, DCR, and PFS compared to MSS-TMB-H, regardless of cut-off (10+ or 20+). Conclusions Regardless of TMB cutoff 10 or 20, MSS-TMB-H does not appear to be a promising predictive biomarker for CPIs compared to MSI-H in GI tumors. Our finding raises the necessity to reconsider TMB-H as a predictive biomarker for CPIs in GI tumors, but further verification is warranted in a larger population. Acknowledgements There are no funding sources Ethics Approval This study was approved by the corresponding Ethical and Independent Institutional Review Board (IRB) from our (Moffitt Cancer Center) institution

With the recent surge in big data analytics for hyper-dimensional data there is a renewed interest in dimensionality reduction techniques for machine learning applications. In order for these methods to improve performance gains and understanding of the underlying data, a proper metric needs to be identified. This step is often overlooked and metrics are typically chosen without consideration of the underlying geometry of the data. In this paper, we present a method for incorporating elastic metrics into the t-distributed Stochastic Neighbor Embedding (t-SNE) and Uniform Manifold Approximation and Projection (UMAP). We apply our method to functional data, which is uniquely characterized by rotations, parameterization, and scale. If these properties are ignored, they can lead to incorrect analysis and poor classification performance. Through our method we demonstrate improved performance on shape identification tasks for three benchmark data sets (MPEG-7, Car data set, and Plane data set of Thankoor), where we achieve 0.77, 0.95, and 1.00 F1 score, respectively.

BACKGROUND:Cancer of unknown primary (CUP) comprises a heterogeneous collection of malignancies that are typically associated with a poor prognosis and a lack of effective treatment options. We retrospectively evaluated the clinical utility of targeted next-generation sequencing (NGS) among CUP patients to assist with diagnosis and identify opportunities for molecularly guided therapy. PATIENTS AND METHODS:Patients with a CUP at Moffitt Cancer Center who underwent NGS between January 1, 2014 and December 31, 2019, were eligible for study inclusion. Next-generation sequencing results were assessed to determine the frequency of clinically actionable molecular alterations, and chart reviews were performed to ascertain the number of patients receiving molecularly guided therapy. RESULTS:Ninety-five CUP patients were identified for analysis. Next-generation sequencing testing identified options for molecularly guided therapy for 55% (n = 52) of patients. Among patients with molecularly guided therapy options, 33% (n = 17) were prescribed a molecularly guided therapy. The median overall survival for those receiving molecularly guided therapy was 23.6 months. Among the evaluable patients, the median duration of treatment for CUP patients (n = 7) receiving molecular-guided therapy as a first-line therapy was 39 weeks. The median duration of treatment for CUP patients (n = 8) treated with molecularly guided therapy in the second- or later-line setting was 13 weeks. Next-generation sequencing results were found to be suggestive of a likely primary tumor type for 15% (n = 14) of patients. CONCLUSION:Next-generation sequencing results enabled the identification of treatment options in a majority of patients and assisted with the identification of a likely primary tumor type in a clinically meaningful subset of patients.

Background Preoperative malnourishment has been consistently associated with poor outcomes after radical cystectomy and other major abdominal surgeries. Most enhanced recovery after surgery (ERAS) studies have examined preoperative nutrition and its relationship to outcomes after gastrointestinal surgery. Although numerous studies have demonstrated the benefits of using an ERAS protocol, this study in unique in comparing 2 ERAS protocols, with and without a nutrition component. Methods A formalized preoperative nutrition protocol (PNP) recommending use of preoperative immunonutrition and carbohydrate drink was introduced in June 2018. A total of 78 consecutive patients who drank both beverages were compared with 92 historical controls. Multivariable logistic regression analyses were sequentially performed to determine if preoperative nutrition was associated with binary outcome variables (30-day complication, infectious complication, and readmission within 30 days). Results The preoperative nutrition group and control group were statistically similar in distribution of age, sex, American Society of Anesthesiologists physical status classification, clinical stage, and body mass index. Return of bowel function was found to occur earlier in the preoperative nutrition group than in the control group (3.12 vs 3.74 days; relative risk, 0.82; CI, 0.73-0.93; P = .0029). Complications within 30 days were similar in both groups (63.6% vs 55.4%; P = 0.36). Infectious complications (42.9% vs 37%; P = .53) and readmission within 30 days (22.1% vs 15.2%; P = .34) were also similar in both groups. Conclusions Use of a PNP including immunonutrition and carbohydrate drink may be associated with earlier return of bowel function after radical cystectomy.

Background Telemedicine for preanesthesia evaluation can decrease access disparities by minimizing commuting, time off work, and lifestyle disruptions from frequent medical visits. We report our experience with the first 120 patients undergoing telemedicine preanesthesia evaluation at Moffitt Cancer Center. Methods This is a retrospective analysis of 120 patients seen via telemedicine for preanesthesia evaluation compared with an in-person cohort meeting telemedicine criteria had it been available. Telemedicine was conducted from our clinic to a patient’s remote location using video conferencing. Clinic criteria were revised to create a tier of eligible patients based on published guidelines and anesthesiologist consensus. Results Day-of-surgery cancellation rate was 1.67% in the telemedicine versus 0% in the in-person cohort. The two telemedicine group cancellations were unrelated to medical workup, and cancellation rate between the groups was not statistically significant ( P = .49). Median round trip distance and time saved by the telemedicine group was 80 miles [range 4; 1180] and 121 minutes [range 16; 1034]. Using the federal mileage rate, the median cost savings was $46 [range $2.30; 678.50] per patient. Patients were similar in gender and race in both groups ( P = .23 and .75, respectively), but the in-person cohort was older and had higher American Society of Anesthesiologists physical status classification ( P = .0003). Conclusions Telemedicine preanesthesia evaluation results in time, distance, and financial savings without increased day-of-surgery cancellations. This is useful in cancer patients who travel significant distances to specialty centers and have a high frequency of health care visits. American Society of Anesthesiologists Physical Status classification and age differences between cohorts indicate possible patient or provider selection bias. Randomized controlled trials will aid in further exploring this technology.

Summary A clearer understanding of the tumor immune microenvironment (TIME) in metastatic clear cell renal cell carcinoma (ccRCC) may help to inform precision treatment strategies. We sought to identify clinically meaningful TIME signatures in ccRCC. We studied tumors from 39 patients with metastatic ccRCC using quantitative multiplexed immunofluorescence and relevant immune marker panels. Cell densities were analyzed in three regions of interest (ROIs): tumor core, tumor–stroma interface and stroma. Patients were stratified into low- and high-marker density groups using median values as thresholds. Log-rank and Cox regression analyses while controlling for clinical variables were used to compare survival outcomes to patterns of immune cell distributions. There were significant associations with increased macrophage (CD68+CD163+CD206+) density and poor outcomes across multiple ROIs in primary and metastatic tumors. In primary tumors, T-bet+ T helper type 1 (Th1) cell density was highest at the tumor–stromal interface (P = 0·0021), and increased co-expression of CD3 and T-bet was associated with improved overall survival (P = 0·015) and survival after immunotherapy (P = 0·014). In metastatic tumor samples, decreased forkhead box protein 3 (FoxP3)+ T regulatory cell density correlated with improved survival after immunotherapy (P = 0·016). Increased macrophage markers and decreased Th1 T cell markers within the TIME correlated with poor overall survival and treatment outcomes. Immune markers such as FoxP3 showed consistent levels across the TIME, whereas others, such as T-bet, demonstrated significant variance across the distinct ROIs. These findings suggest that TIME profiling outside the tumor core may identify clinically relevant associations for patients with metastatic ccRCC.

The Citrus tristeza virus (CTV) is one of the most economically devastating citrus diseases worldwide. The spread of CTV in eastern Spain was studied by Gottwald et al. with the goal of determining the spatio-temporal mechanisms of spread. Since the subjects in this study are individual trees, it is natural to think of infections as Bernoulli trials. This approach is difficult however, due to the spatial and temporal dependence of the observations. Consequently, a system of ordinary differential equations (ODE) was used to model the probabilities of infection as well as the spatial and temporal dependence. Given the parameters in the ODE, the probabilities of infection are treated as conditionally independent. Using the conditional independence we then specify the joint likelihood function as a Poisson binomial distribution. For the purpose of model selection and hypothesis testing we, employed accumulated prediction error (APE) which has connections to both Bayesian and frequentist frameworks. We demonstrated the robustness of our method in accounting for spatio-temporal dependencies in the data by accurately predicting the spatial distribution of the disease through Join Counts.

Introduction Mycophenolate mofetil (MMF) dose/kg can influence clinical outcomes of allogeneic hematopoietic cell transplantation (HCT) including acute graft-versus-host disease (GVHD) (Harnicar BBMT 2015; Bejanyan BBMT 2018). When used for GVHD prophylaxis in conjunction with post-transplant cyclophosphamide (PTCy), MMF is generally dosed at 15 mg/kg three times daily (capped at a maximum of 3 g/day) and combined with either tacrolimus or sirolimus. Thus, many adults receive the maximum dose and a variable dose/kg of MMF. Objectives We sought to investigate the impact of MMF dose/kg on clinical outcomes of HCT in the setting of PTCy-based GVHD prophylaxis, where the effect of MMF dose/kg is unknown. Methods Included are 112 consecutive adult patients with hematologic malignancies who received myeloablative (n=70) or reduced intensity (n=42) HCT at the Moffitt Cancer Center between 2012-2018 with T-cell replete peripheral blood stem cells (n=94) or bone marrow (n=18). GVHD prophylaxis consisted of PTCy/MMF and either tacrolimus (n=57) or sirolimus (n=55). MMF dose relative to patient actual body weight (mg/kg/day), was stratified by tertiles into low ( 40 mg/kg/day, n=36). Results The median age at transplant was 57.8 (range: 21-75) years. Median follow-up was 12 months (range: 0.3-39). Donors were haploidentical related (n=74), mismatched unrelated (n=26), and matched related/unrelated (n=12). MMF dose groups had similar characteristics, except the low dose group having a higher proportion of males (p Conclusion In patients receiving PTCy based GVHD prophylaxis, MMF dose/kg had no significant effect on GVHD. However, lower MMF dose/kg was associated with worse DFS. This study suggests that

Comparison of ribonucleic acid (RNA) molecules is important for revealing their evolutionary relationships, predicting their functions and predicting their structures. Many methods have been developed for comparing RNAs using either sequence or three-dimensional (3D) structure (backbone geometry) information. Sequences and 3D structures contain non-overlapping sets of information that both determine RNA functions. When comparing RNA 3D structures, both types of information need to be taken into account. However, few methods compare RNA structures using both sequence and 3D structure information. Recently, we have developed a new method based on elastic shape analysis (ESA) that compares RNA molecules by combining both sequence and 3D structure information. ESA treats RNA structures as 3D curves with sequence information encoded on additional coordinates so that the alignment can be performed in the joint sequence-structure space. The similarity between two RNA molecules is quantified by a formal distance, geodesic distance. In this study, we implement a web server for the method, called RASS, to make it publicly available to research community. The web server is located at http://cloud.stat.fsu.edu/RASS/.