We reviewed the clinical characteristics of male patients with MEN2B associated medullary thyroid cancer (MTC) with prostate lesions and analyzed available imaging and pathology data.Patients were enrolled on a National Cancer Institute (NCI) natural history protocol (NCT01660984) and a University of Texas (UT) MD Anderson Cancer Center study (DR09-0507).Thirty-six male patients (median age of MEN2B diagnosis 12.0 years, range 1.3-25.6) with the RET p.Met918Thr germline mutation were identified.Pelvic imaging was available for 28/36 (78%) patients. Prostate lesions and/or calcifications were noted in eight patients (28.6%). Lesions were identified at a median age of 20.5 years (range 13-30.5 years). Biopsies in three patients with prostate neoplasms were histologically indistinguishable from primary MTC. Two of these patients did not have other sites of distant disease at the time of prostate biopsy. Prostate lesions in males with MEN2B and MTC are more frequent than previously described. Lesions were identified in patients as young as 14 years and often contained calcifications. The prostate is a more common site of disease in patients with MEN2B than previously appreciated and may represent a distant site of metastatic MTC metastasis or a second primary calcitonin-secreting neuroendocrine tumor. Imaging of the prostate should be considered in post-pubertal MEN2B males, especially in case of unexplained increase in calcitonin or onset of urinary tract symptoms.

Background/Objectives: Prostate cancer is the second most frequent cancer among men and the 5th leading cause of cancer death among men worldwide. Identification of urine-derived biomarkers, such as exosomal miRNAs, in liquid biopsies for prostate cancer could be very beneficial for screening and active surveillance. Methods: Urine was collected from 42 patients with biopsy-proven evidence of prostate cancer and exosomes were extracted. Transcriptomic analysis was performed on the urine-derived exosomal miRNA and compared to the urine-derived exosomal miRNA profiles from 10 normal control donors and 15 von Hippel-Lindau (VHL) syndrome patients with clear cell renal cell carcinoma (ccRCC). Results: Urine-derived exosomal miRNA profiles of prostate patients were significantly different from normal control individuals. Significantly increased expression of miR-122-5p and decreased expression of miR-125-5p and miR-16-5p were observed in the urine-derived exosomes from prostate cancer patients. Significant upregulation of miR-30a-5p and downregulation of miR-320-5p, miR-320b, and miR-320c were observed in the urine-derived exosomes from both prostate cancer patients and VHL patients with ccRCC, indicating these miRNAs could be non-specific markers of urological cancer. Increased expression of miR-10a-5p and miR-30e-5p or miR-532-5p and miR-206 correlated with the presence of either extracapsular or perineural invasion, respectively. Conclusions: This study highlights the potential for urine-derived exosomal miRNA profiles to identify the presence of prostate cancer and predict clinical features, additionally showing that miRNA signals could be non-specific markers of urologic cancer types. Further validation studies are necessary to demonstrate the utility of urine-derived exosomal miRNA profiles as biomarkers for diagnosis or prognosis in prostate cancer.

Correlations between pre- and posttreatment MRI tumor volumes

315 Background: PCa has substantial inherited predisposition and certain germline variants like BRCA1/2 , ATM , HOXB13 , and DNA mismatch repair (MMR) genes are associated with an increased risk of PCa. This study follows men without a diagnosis of PCa, with known germline pathogenic/likely pathogenic variant (PV) in BRCA1/2 , DNA MMR genes associated with Lynch syndrome ( MLH1/PMS2 , MSH2 / MSH6 , EPCAM ), HOXB13 , ATM , CHEK2 , PALB2 , TP53 , NBN , RAD51C/D , BRIP1 , and FANCA-FANCM (NCT03805919). Methods: Up to 500 men, ages 30-75 years old (y/o) with a documented PV will enroll. Men undergo biennial mpMRI and annual PSA. Indication for prostate biopsy includes clinical, PSA, and/or MRI findings. Patients are followed at 12-month intervals to determine PSA, PCa diagnosis, and disease/survival status until death. Results: To date, 378 evaluable patients have enrolled: 353 (93%) Caucasian, 11 (3%) Hispanic, 7 (2%) Asian, 4 (1%) African-American, 2 (1%) bi-ethnic. 1 Indian-Asian. Median age was 47 y/o. The most common PV were: 180 (48%) BRCA2 , 94 (25%) BRCA1 , 22 (6%) CHEK2 , and 18 (5%) ATM . PVs in MLH1 / PMS2 , MSH2 / MSH6 , PALB2 , HOXB13 , TP53 , BRIP1 , RAD51D , EPCAM , NBN , and FANCA are <4%. Eight patients carried more than one PV. A total of 782 MRIs were performed: 378 baselines, 228 at year 2, 121 at year 4, 30 at year 6, and 25 clinical. Indication for biopsy was present in 89 (24%) patients with 37 (10%) diagnosed with PCa. Of the 90 biopsies indicated, 1 refused and withdrew and 3 are still pending. Of those with PCa, 22 had BRCA1/2 PVs, 5 had MMR PVs, and 10 had non-MMR PVs. Median age at diagnosis was 62. 24 patients were diagnosed with Grade Group (GG) 1, 8 patients with GG2, 1 patient with GG3, 3 patients with GG4, 1 patient with GG5. 21 opted for active surveillance (AS), 10 opted for prostatectomy, 6 opted for radiation therapy. 5 patients on AS converted to definitive treatment after progression was noted at 1-year follow-up. Conclusions: mpMRI screening in men with PVs is feasible and can be used for early diagnosis, PCa monitoring, and facilitates PCa diagnosis at PSA levels below conventional thresholds. Correlative studies including cfDNA, PBMCs and PRS, are ongoing. Clinical trial information: NCT03805919 . Indication for biopsy. BiopsyPositive BiopsyNegative Biopsy Pending, Refused Total = 89 PSA WNL PSA Elevated PSA WNL PSA Elevated PSA WNL PSA Elevated PIRADS 1 13 0 4 0 9 0 0 PIRADS 2 15 1 3 5 6 0 0 PIRADS 3 23 5 1 11 3 2 1 PIRADS 4 34 10 9 13 2 0 0 PIRADS 5 4 1 3 0 0 0 0

Prostate cancer (PCa) is the second most common cancer and cause of cancer death in American men. Existing risk prediction methods have limited accuracy and reproducibility, resulting in difficulty in predicting disease severity. We demonstrate the development and external validation of an automated multimodal artificial intelligence algorithm using biparametric MRI (bpMRI) and clinical covariates for predicting biochemical recurrence (BCR) after radical prostatectomy (RP) in PCa patients. Development cohort included 80% of patients from center 1 (n = 240) who underwent prostate MRI prior to RP between January 2008 and December 2018 with a minimum of two years of follow-up after RP. Test cohort included the remaining 20% of center 1 patients (n = 71), and the external validation cohort from center 2 (n = 168). Center 2 patients included those who underwent prostate MRI and RP between January 2015 and December 2024 with a minimum of two years of follow-up. Clinical comparisons were CAPRA-S (center 1) and ISUP grade group from post-RP biopsy (center 2). Models developed were a clinical model (M0), an automated clinical model (M1), a radiomics model (M2), and a multimodal model (M3). Clinical variables (M0) included PSA, age, primary Gleason, and ISUP grade group. Automated clinical variables (M1 and M3) included PSA and age. Radiomics features (M2 and M3) were extracted from bpMRI using a lesion detection algorithm. Accuracy, sensitivity, specificity, and AUC were calculated, and log-rank tests compared BCR-free survival to assess the models' ability to discriminate relative to clinical standards. Intermediate-risk groups were also assessed. The multimodal model (M3) had the highest AUC across test sets (combined: 0.71; center 1: 0.70; center 2: 0.75) and was the only model to significantly differentiate BCR-free survival outcomes in intermediate-risk groups across both centers (p < 0.05). This automated multimodal model leveraging radiomics and clinical covariates can predict BCR after RP, approaches clinical gold standards, and may enhance imaging-based prognostication following further validation.

ABSTRACT Background Biochemical recurrence (BCR) occurs in 20–40% of men after radical prostatectomy. Existing postoperative recurrence risk tools based on PSA and pathology are clinically useful but show only moderate and variable discrimination, highlighting the need for biomarkers that improve risk stratification and consequent treatment decisions. We hypothesized that the prostate microenvironment, including both the tumor and non-cancerous adjacent tissue, may contain prognostic features associated with adverse postoperative PSA outcomes. Methods We assembled a cohort of matched tumor-adjacent benign and tumor prostate tissue from 243 men across three institutions to establish a discovery cohort (n=123; 43 postoperative PSA events, 35%) and validation cohort (n=120; 46 events, 38%). For primary binary analyses, a postoperative PSA event included BCR, defined as two consecutive postoperative PSA values ≥0.2 ng/mL, or PSA persistence. We performed RNA sequencing of matched tumor-adjacent benign and tumor tissues, quantified immune signatures, and developed an integrated model combining the adjacent-tissue B-cell signature, preoperative PSA, and radical prostatectomy Gleason score (“BRIGADE”). CAPRA-S-adjusted Cox analyses excluding recurrence-time-0 cases evaluated time to BCR, and CD19 multiplex immunofluorescence provided tissue-level confirmation (n=10). Results In prostatectomy specimens, tumors from patients without a postoperative PSA event were enriched for B-cell transcriptional programs, whereas tumors from event-positive patients showed elevated proliferation signatures. B-cell-related transcriptional programs were correlated between tumor and adjacent tissue. Tumor-adjacent benign B-cell scores were higher in no-event cases and discriminated postoperative PSA-event status in PCBN discovery (AUC 0.63) and BM validation (AUC 0.81) cohorts, outperforming numerous other immune-related signatures. In CAPRA-S-adjusted Cox sensitivity analyses excluding recurrence-time-0 cases, higher adjacent-tissue B-cell activity was associated with reduced recurrence risk in PCBN (HR 0.42, 95% CI 0.19–0.94; BH-adjusted p=0.035) and BM (HR 0.54, 95% CI 0.30–0.95; BH-adjusted p=0.034). Tissue-based validation showed that CD19⁺ B-cell density in adjacent benign tissue was higher in no-event than event-positive patients (median 0.1145 vs 0.0471; p=0.008). BRIGADE achieved an AUC of 0.68 in cross-validation and 0.83 in independent validation, compared to AUCs of 0.54–0.63 and 0.44–0.78 for the tested clinical predictors, respectively. At the fixed classification threshold, the validation-cohort odds ratio for BRIGADE was 2.75. The adjacent B-cell score remained associated with lower odds of a postoperative PSA event after adjustment for PSA and Gleason score. Conclusions B-cell infiltration in tumor-adjacent benign prostate tissue may complement existing clinicopathologic models for stratifying adverse postoperative PSA outcomes and subsequent BCR after radical prostatectomy. The transcriptomic signal was recapitulated by CD19-based tissue staining, supporting further development of a pathology-based assay.

Radiomic features of radiation response

OBJECTIVE:To identify molecular features associated with earlier progression to definitive therapy amongst patients with localized prostate cancer (PCa) managed on active surveillance (AS). METHODS:We performed a retrospective pilot study of 7 patients with low- to intermediate-risk PCa undergoing serial multiparametric MRI (mpMRI)-targeted biopsies of the same lesion while on AS, who all proceeded to definitive therapy. Time-to-treatment (TTT) was defined as years from first biopsy on AS to definitive therapy. Laser-capture microdissection was used to separate tumor epithelium, benign glands, high-grade prostatic intraepithelial neoplasia, and stroma in each biopsy specimen. DNA from the tumor and matched benign tissue underwent whole-exome sequencing, and RNA from all compartments underwent whole-transcriptome sequencing. Somatic mutations and copy-number alterations were compared across serial biopsies and used to reconstruct phylogenies and quantify clonal complexity. RESULTS:Tumors exhibited substantial intratumoral heterogeneity, and in 3 of 6 paired cases, serial mpMRI-targeted biopsies showed discordant somatic profiles consistent with sampling distinct major clones over time. By contrast, no single gene-level alteration, and few large-scale chromosomal events, were associated with TTT. High clonal complexity, defined as ≥3 subclones, was associated with significantly shorter TTT than low complexity (median 1.9 vs 7.2 years; P = .0082). Exploratory pathway analyses of individual tissue components suggested TTT-associated differences in inflammatory signaling and stromal-epithelial cross-talk. CONCLUSION:In this small, hypothesis-generating cohort, clonal complexity was more closely associated with earlier definitive therapy than individual genomic alterations. Larger prospective studies are needed to validate whether multiomic measures of clonal architecture can improve AS risk stratification.

RATIONALE AND OBJECTIVES:To develop a pathology-derived radiomics signature for detecting clinically significant prostate cancer (csPCa) and to evaluate its performance using lesion diameter-based simplified segmentations. MATERIALS AND METHODS:In this retrospective single-center study, 175 participants (120 radical prostatectomy cases; 55 controls) underwent biparametric MRI during 2013-2022. Whole-mount histopathology was registered to MRI using a patient-specific, mold-based 3D pipeline to generate lesion-level ground truth. Six radiologists from different institutions marked lesion diameters per Prostate Imaging Reporting & Data System (PI-RADS) v2.1, blinded to pathology; automated circular segmentations were generated from these measurements and expanded (±1 slice). Features (PyRadiomics) were preprocessed, filtered, and benchmarked via nested cross-validation. Recursive feature elimination produced a 10-feature pathology-derived radiomics signature. Models (Signature, prostate-specific antigen density [PSAD], PI-RADS, and their combinations) were trained/evaluated using a soft-voting ensemble (logistic regression, random forest, and XGBoost) with patient-level grouping; thresholds were optimized using Youden's J. DeLong's and McNemar's tests were used for model comparisons. RESULTS:PSAD+Signature achieved 0.75 area under the curve (AUC) and 68% (211/312) accuracy. PI-RADS+PSAD achieved 0.77 AUC and 74% (230/312) accuracy. Signature-only achieved 0.66 AUC and 62% (194/312) accuracy. A higher AUC for PSAD+Signature versus Signature (|ΔAUC|=0.093, p=0.012) and for the tripartite model versus Signature (|ΔAUC| = 0.11, p=0.007) was found. PSAD+Signature and PI-RADS+PSAD had similar accuracy (p=0.06). CONCLUSION:A histopathology-trained radiomics signature demonstrated moderate standalone performance for lesion-level csPCa detection. When combined with PSAD, diagnostic performance improved and approached that of PI-RADS+PSAD, which achieved the highest absolute accuracy. The PSAD+Signature framework offers a simplified, spatially localized approach that may complement existing PI-RADS-based assessment while maintaining low implementation complexity.

PURPOSE:Prostate cancer (PCa) is the second most common cancer and cause of cancer deaths among American men. Existing risk prediction methods have limited accuracy and reproducibility, resulting in difficulty in predicting treatment outcomes. We demonstrate the development and external validation of an automated multimodal artificial intelligence (AI) algorithm using biparametric magnetic resonance imaging (bpMRI) and clinical covariates for predicting biochemical recurrence (BCR) after radical prostatectomy (RP) in patients with PCa. METHODS:The development cohort included 80% of patients from center 1 (n = 240) who underwent prostate MRI prior to RP between January 2008 and December 2018, with a minimum of 2 years of follow-up after RP. The test cohort included the remaining 20% of center 1 patients (n = 71) and an external validation cohort from center 2 (n = 168). Center 2 patients included those who underwent prostate MRI and RP between January 2015 and January 2024, with a minimum of 2 years of follow-up. Clinical comparisons were made using the Cancer of the Prostate Risk Assessment Postsurgical (center 1) and International Society of Urological Pathology Gleason Grade Group (ISUP GGG) scoring systems from post-RP pathology (center 2). The models developed were as follows: clinical (M0), automated clinical (M1), radiomics (M2), and a multimodal model (M3). Clinical variables (M0) included prostate-specific antigen (PSA), age, primary Gleason, and ISUP GGG. Automated clinical variables (M1 and M3) included PSA and age. Radiomic features (M2 and M3) were extracted from bpMRI using a lesion detection AI model. Accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were calculated, and log-rank tests compared BCR-free survival to assess the models' ability to discriminate relative to clinical standards. Intermediate-risk groups were also assessed. RESULTS:The multimodal model (M3) had the highest AUC across test sets (combined: 0.71; center 1: 0.70; center 2: 0.75). This was the only model that significantly differentiated BCR-free survival outcomes in intermediate-risk groups across both centers (P < 0.05). CONCLUSION:This automated multimodal model leveraging radiomics and clinical covariates can predict BCR after RP, approaching clinical gold standards, and may enhance imaging-based prognostication following further validation. CLINICAL SIGNIFICANCE:Given that this model demonstrated the potential to outperform pre-surgical and post-surgical clinical gold standards in an external cohort's intermediate-risk patient subgroup (for whom it is more challenging to predict disease trajectory), this model may contribute to enhanced personalized care in PCa after further validation.

e17123 Background: The most recent guidelines from the European Urology Association (EUA) have recommended implementation of perilesional sampling for prostate biopsies in addition to lesion center sampling during targeted biopsies (TBx). In this study we investigate the spatial distribution of high risk prostate cancer patterns on whole mount digital pathology slides. Methods: 1108 whole mount pathology slides from 316 men were included in this study. Digital slides were annotated by a GU pathologist to reflect intratumoral heterogeneity of morphology patterns. Annotated histology patterns were categorized into three different groups: low (Gleason Grade (GG)1), intermediate (GG2, GG3), and high risk (GG4, GG5, Cribriform). The spatial distribution of patterns within tumor foci was evaluated with respect to 1 and 3mm defined interior edges. The location of a pattern was defined as the ratio from 0 to 1 of the amount of a pattern found within the defined edge and size of the pattern, with higher values indicating pattern proximity to the edge. Tumor regions too small to define an edge sperate from a center and one-pattern regions were excluded. A linear mixed effects model summarized the effects of patient GG, tumor foci size, tumor pattern size, and pattern risk on the edge ratio for 1 and 3mm edges, calculated as pixel ratio with respect to low risk patterns. Results: Median edge ratio values for 1 and 3mm edges were 0.70 and 1.0 for low risk patterns, 0.30 and 0.72 for intermediate risk patterns, and 0.17 and 0.69 for high risk patterns, respectively. Intermediate (β=-0.17, p<0.001) and high risk patterns (β=-0.23, p<0.001) were less localized to both interior edges. Spatial distributions at both edges differed for all paired risk groups. The size of histology patterns (β=-0.06, p<0.001) and tumor foci (β=-0.10, p<0.001) were also associated with less localization towards both interior edges. At the 1mm edge, GG5 patients (n=27) had an increase (β=0.11, p<0.001) in incidence towards the edge, but was fewer than GG4 (n=37) and GG3 (n=53). Conclusions: Our results reveal that high risk patterns are more commonly located at the center of tumors compared to low risk patterns. These results highlight the importance of lesion center sampling approach during TBx to identify more aggressive GG patterns. The effects of patient GG, tumor foci size, pattern size, and pattern risk on the edge ratio for 1 and 3mm edges, calculated as pixel ratio ~ (1|MRN/tumor foci) + high risk pattern + scale (pixels pattern) + scale (pixels tumor foci) + patient GG are summarized. 1mm edge 3 mm edge Predictors β estimates p β estimates p intermediate risk pattern -0.17 <0.001 -0.03 0.102 high risk pattern -0.23 <0.001 -0.09 <0.001 pixels pattern -0.06 <0.001 -0.07 <0.001 pixels tumor foci -0.10 <0.001 -0.09 <0.001 patient GG3 0.01 0.806 -0.00 0.886 patient GG4 0.02 0.309 -0.04 0.109 patient GG5 0.11 <0.001 0.04 0.133

BACKGROUND:In the modern magnetic resonance imaging (MRI) era, prostate-specific antigen (PSA) density (PSAD) derived from multiparametric MRI (mpMRI) has become an accessible biomarker for risk stratification in localized prostate cancer. Although widely used, the optimal PSAD threshold for predicting adverse pathology and long-term oncologic outcomes after radical prostatectomy remains uncertain. OBJECTIVE:To determine the prognostic utility and optimal cutoff of MRI-based PSAD in predicting adverse pathology and oncologic outcomes among men undergoing robotic-assisted laparoscopic radical prostatectomy (RALP). DESIGN, SETTING, AND PARTICIPANTS:This study analyzed a prospectively maintained MRI-era cohort of 788 consecutive patients who underwent RALP between 2006 and 2023. All patients had preoperative mpMRI with volumetric assessment, pathology data, and long-term follow-up for biochemical and metastatic progression. The primary endpoint was adverse pathology, defined as ≥pT3 disease, Grade Group ≥4, seminal vesicle invasion (SVI), lymph-node positivity, or positive surgical margin (PSM). Secondary outcomes included biochemical recurrence-free survival (BRFS). RESULTS:Median age was 62.5 yr, median was PSA 6.7 ng/ml, median prostate volume was 40.0 ml, and median MRI-based PSAD was 0.16 ng/ml/cm3. Adverse pathology occurred in 43% of patients, with 15.7% harboring PSMs, 7.4% SVI, and 4.7% nodal involvement. On multivariable analysis, MRI-based PSAD independently predicted adverse pathology (odds ratio [OR] 15.633, p < 0.001), SVI (OR 2.614, p = 0.040), PSM (OR 3.650, p = 0.028), and PSA persistence (OR 4.463, p = 0.004). Decision-curve analysis demonstrated that PSAD ≥ 0.10 ng/ml/cm3 provided the greatest overall net benefit in Prostate Imaging-Reporting and Data System (PI-RADS) ≥3 for predicting adverse pathology, while PSAD ≥0.15 ng/ml/cm3 achieved better specificity and performed best in men with PI-RADS ≤2. Kaplan-Meier analyses showed significantly worse BRFS for PSAD ≥0.15 ng/ml/cm3 versus <0.15 ng/ml/cm3 in PI-RADS ≤2 (p = 0.002) and for PSAD ≥0.10 ng/ml/cm3 versus <0.10 ng/ml/cm3 in PI-RADS ≥3 (p = 0.001). CONCLUSIONS:MRI-based PSAD is a strong, independent predictor of adverse pathology and oncologic outcomes after RALP in the MRI era. A threshold of 0.15 ng/ml/cm3 offers the most balanced clinical utility in men with PI-RADS ≤2 lesion, whereas 0.10 ng/ml/cm3 may be more appropriate in men with PI-RADS ≥3 lesion.

BACKGROUND:Biochemical recurrence (BCR) following radical prostatectomy (RP) occurs in 20%-40% of patients with localized prostate cancer. Neoadjuvant PROSTVAC has been shown to enhance T-cell infiltration into the tumor immune microenvironment, but whether these immunologic changes translate into improved long-term oncologic outcomes remains unknown. PATIENTS AND METHODS:This secondary analysis evaluated 26 patients from a Phase II neoadjuvant PROSTVAC trial who underwent RP. Patients received recombinant vaccinia-PSA-TRICOM priming followed by three fowlpox-PSA-TRICOM boosts before surgery. Tumor immune responses (CD4 + /CD8 + T-cell infiltration) and peripheral antigen-specific T-cell responses to PSA, MUC-1, and brachyury were previously assessed. Primary outcomes included BCR-free survival (BCR-FS) and metastatic progression at extended follow-up. RESULTS:From May 2014 to June 2017, 26 patients were enrolled, received neoadjuvant PROSTVAC, and subsequently underwent RP. Eighteen (69.2%) patients had NCCN® unfavorable-intermediate, high, or very high-risk disease at baseline. At a median follow-up of 8.9 years (range 7.0-10.1), BCR occurred in six patients (23.1%), and none developed metastatic disease. Comparing pre- and post-PROSTVAC peripheral antigen-specific T-cell responses, patients with PSA-specific immune responses had 100% 8-year BCR-FS, compared with 70.0% in nonresponders (p = 0.247). Those with any tumor-associated antigen response demonstrated 85.7% versus 66.7% 8-year BCR-FS (p = 0.149). Patients with CD4+ or CD8 + T-cell responses at the tumor site or invasive margin had 86.2% versus 58.9% 8-year BCR-FS (p = 0.257). CONCLUSION:Neoadjuvant PROSTVAC may have oncologic activity in localized prostate cancer, as we observed a longer BCR-FS in immune responders and no metastatic progression in a small cohort. Further investigation of the biological mechanism by which therapeutic immunotherapy may establish durable antitumor immunity in localized prostate cancer is warranted.

Gene list of a prostate cancer-specific TGF-β activity signature

Despite the apparent simplicity of the motor action involved during percutaneous needle procedures, manipulating the tool's direction becomes challenging when clinicians cannot directly visualize internal anatomy and must rely on ultrasound images, which increase cognitive demand. Augmented reality (AR) offers the promise to assist with these tasks by providing pertinent visual information in the clinician's field of view. However, simply providing visual information that ignores meaningful visual cues can complicate depth perception and spatial understanding. In this work, we introduce and evaluate three visualization techniques for needle alignment developed during design sessions with medical experts: a localized focus-and-context window, a color-based proximity encoding, and an explicit trajectory overlay. These techniques were evaluated in a user study (n=26) including clinical experts (n=7) using a prostate-biopsy-inspired phantom. Results from this study suggest that cue effects depended on the surrounding cue configuration and user expertise. For novices, explicit trajectory overlay improved targeting accuracy and reduced retreat behavior, but its effect on completion time varied across cue configurations, with slower performance when the overlay was presented alone. For experts, the focus-and-context window reduced completion time and retreat events, while color-based proximity overlay improved completion time. Subjectively, color cues were often perceived as helpful even when their effects on accuracy were not consistent. These results suggest that AR guidance strategies for percutaneous interventions should consider user expertise and visual context.

Introduction Integrating large language models (LLMs) into healthcare is set to transform medical research. Most clinical research relies on data manually extracted by data managers, a laborious and time-consuming process. To streamline such tasks, the National Institutes of Health Integrated Data Analysis Platform (NIDAP) Text Extraction Program (NTEP) was developed. This artificial intelligence data aggregation platform, powered by LLMs, can output a collection of data within seconds after a prompt engineering process by the clinician. In this study, we aim to compare the accuracy of data extracted by NTEP with data that was manually extracted by NIH data managers in patients with prostate cancer enrolled in our institution's prospective trial. Methods We conducted a comparative analysis between datasets extracted by data managers and NTEP. Both were tasked to extract data for four MRI-related variables for patients enrolled in the prostate cancer natural history trial (NCT02594202): prostate volume, PSA density, number of lesions, and PI-RADS score. Custom-built LLM prompts were built by urologists using GPT-4 prompts aimed to extract the data directly from electronic medical record (EMR) documents. Both datasets were then subject to minor processing and formatting to allow for comparison between extraction methods. Prostate volumes were rounded to the appropriated absolute value, PSA density was rounded to three decimals places, and only the highest PI-RADS lesion reported by data managers was evaluated. Statistical analysis was performed with SPSS 29.0 to evaluate the correlation between pair observations in continuous variables via a Spearman's rho, and to quantify the level of agreement between categorical variables, a Cohen's kappa was performed. Results A total of 1728 MRIs from 1289 patients were evaluated. In comparing the datasets extracted by NIDAP and the data managers, we found that agreement between values occurred 1598 times (92.5%) for prostate volume, 1705 times (98.7%) for PSA density, 1221 times (70.7%) for number of lesions, and 1577 times (91.3%) for PI-RADS score. In reports that had pair observations, both NIDAP and data managers results appeared highly concordant, however, the results between both groups differed from 0.5% to 6.8%. There were also cases where the datasets were missing data entirely; notably, for the number of lesions on MRI, the data managers did not report data in 488 (28.2%) instances. (Table 1) Conclusions NTEP is a useful tool to facilitate data extraction from EMRs. Although there is a high concordance when data was reported by both NIDAP and data managers, NIDAP was able to extract more information, leading to fewer missing variables. Future research should involve larger cohorts to validate the platform's scalability and efficiency compared to traditional manual extraction methods, and data quality extracted by NTEP should be further assessed. We anticipate that the integration of LLMs will significantly enhance and transform the data extraction process.

PURPOSE:To develop a multimodal deep learning-based AI algorithm and investigate its ability to predict BCR of PCa after radical prostatectomy (RP) using MRI and clinical data. METHODS:PCa patients (n = 311) underwent prostate MRI prior to RP between January 2008 and December 2018. For each patient, CAPRA-S was calculated. Quantitative imaging features were extracted using methods developed in a previous study. Test set results were assessed independently for each model in the study, using cross-validation of the training set to tune hyperparameters and select features. DeLong's test compared AUROC curve values, and log-rank tests compared BCR-free survival curves. RESULTS:Across all patients, the AUROC of the automated multimodal model was 0.74, compared to 0.66 for CAPRA-S. This model had the highest sensitivity at 75 %, with CAPRA-S at 37 %. BCR-free survival curves for the test set were generated for each model. Log-rank tests indicated each model differentiated between patient outcomes (p < 0.05). The automated multimodal model was the only model with p < 0.01. Focusing on intermediate risk patients (CAPRA-S scores 3-5), this automated model was the only model which maintained the ability to differentiate between outcomes (p < 0.01), while all other models and CAPRA-S failed to differentiate intermediate risk BCR outcomes (p > 0.05). CONCLUSION:Development of a multimodal model using quantitative imaging features and clinical covariates revealed that an automated multimodal AI approach most effectively predicts BCR in PCa patients. Based on AUROC and the ability to differentiate between BCR-free survival outcomes with statistical significance in intermediate risk patients, this model outperforms the gold standard postsurgical CAPRA-S risk scores.

With the overall technological diagnostic improvement of prostate cancer (PCa), focal therapy has emerged as a promising approach for the treatment of localized PCa, offering in a selected group of patients an intermediate option between active surveillance and radical interventions. The primary goal of focal therapy is to avoid local progression and metastatic disease, and decrease the morbidity associated with whole gland therapy. Selection of energy source and approach depends on the index lesion(s) location, size, prostate anatomy, surrounding structures, overall clinical characteristics, patient expectations, and surgeon experience. Further long-term prospective data assessing the outcomes of focal therapy are still required.

Augmented reality (AR) technologies enable the superimposition of imaging upon a patient in real time with three dimensional instrument tracking during procedures. We sought to demonstrate the feasibility of using an AR system (XR90, MediView XR Inc., Cleveland, OH) to fuse a pelvic multi-parametric magnetic resonance image segmentation with ultrasound to perform a non-rectal, fully trans-perineal (FTP), AR-assisted prostate biopsy. AR-assisted biopsy results were congruent with standard fusion biopsy results, showing benign prostate tissue. No adverse events occurred. Limitations include the current workflow and reliance on a non-specialized ultrasound probe.