Comparing Patient-reported Functional Outcomes After Radical Prostatectomy in Historical and Contemporary Practice.

PURPOSE: Modifications to surgical technique, particularly the widespread adoption of robotic surgery, have been proposed to improve functional recovery after prostate cancer surgery. However, rigorous comparison of men in historical vs contemporary practice to evaluate the cumulative effect of these changes on urinary and sexual function after radical prostatectomy is lacking. MATERIALS AND METHODS: We compared prospectively collected patient-reported urinary and sexual function from historical (PROSTQA [Prostate Cancer Outcomes and Satisfaction With Treatment Quality Assessment study], n=235) and contemporary (MUSIC-PRO [Michigan Urological Surgery Improvement Collaborative Patient Reported Outcome] registry, n=1,215) cohorts at the University of Michigan to understand whether modern techniques have resulted in functional improvements for men undergoing prostate cancer surgery. RESULTS: We found significant differences in baseline function, with better urinary (median [IQR]; 100 [93.8-100] vs 93.8 [85.5-100], P < .001) and sexual scores (median [IQR]; 83.3 [66.7-100] vs 74.4 [44.2-87.5], P < .001) prior to treatment in PROSTQA compared to MUSIC-PRO patients, respectively. There was no statistically significant difference in the pattern of urinary incontinence recovery after surgery from 6-24 months between groups (P = .14). However, men in the contemporary MUSIC-PRO group did have significantly better recovery of sexual function compared to men in the historical PROSTQA group (P < .0001). Further, we found that contemporary practice consists of men with more unfavorable demographic and clinical characteristics compared to historical practice. CONCLUSIONS: Our results demonstrate that the widespread alterations in prostate cancer surgery over the past 2 decades have yielded improvements in sexual, but not urinary, function recovery.

Active surveillance for prostate cancer: selection criteria, guidelines, and outcomes.

INTRODUCTION: Active surveillance (AS) has been widely adopted for the management of men with low-risk prostate cancer. However, there is still a lack of consensus surrounding the optimal approach for monitoring men in AS protocols. While conservative management aims to reduce the burden of invasive testing without compromising oncological safety, inadequate assessment can result in misclassification and unintended over- or undertreatment, leading to increased patient morbidity, cost, and undue risk. No universally accepted AS protocol exists, although numerous strategies have been developed in an attempt to optimize the management of clinically localized disease. Variability in selection criteria, reclassification, triggers for definitive treatment, and follow-up exists between guidelines and institutions for AS. In this review, we summarize the landscape of AS by providing an overview of the existing AS protocols, guidelines, and their published outcomes. METHODS: A comprehensive electronic search was performed to identify representative studies and guidelines pertaining to AS selection criteria and outcomes. CONCLUSION: While AS is a safe and increasingly utilized treatment modality for lower-risk forms of PCa, ongoing research is needed to optimize patient selection as well as surveillance protocols along with improved implementation across practices. Further, assessment of companion risk assessment tools, such as mpMRI and tissue-based biomarkers, is also needed and will require rigorous prospective study.

Multigene Profiling of Circulating Tumor Cells (CTCs) for Prognostic Assessment in Treatment-Naïve Metastatic Hormone-Sensitive Prostate Cancer (mHSPC).

The substantial biological heterogeneity of metastatic prostate cancer has hindered the development of personalized therapeutic approaches. Therefore, it is difficult to predict the course of metastatic hormone-sensitive prostate cancer (mHSPC), with some men remaining on first-line androgen deprivation therapy (ADT) for several years while others progress more rapidly. Improving our ability to risk-stratify patients would allow for the optimization of systemic therapies and support the development of stratified prospective clinical trials focused on patients likely to have the greatest potential benefit. Here, we applied a liquid biopsy approach to identify clinically relevant, blood-based prognostic biomarkers in patients with mHSPC. Gene expression indicating the presence of CTCs was greater in CHAARTED high-volume (HV) patients (52% CTChigh) than in low-volume (LV) patients (23% CTChigh; * p = 0.03). HV disease (p = 0.005, q = 0.033) and CTC presence at baseline prior to treatment initiation (p = 0.008, q = 0.033) were found to be independently associated with the risk of nonresponse at 7 months. The pooled gene expression from CTCs of pre-ADT samples found AR, DSG2, KLK3, MDK, and PCA3 as genes predictive of nonresponse. These observations support the utility of liquid biomarker approaches to identify patients with poor initial response. This approach could facilitate more precise treatment intensification in the highest risk patients.

Analysis of geminivirus AL2 and L2 proteins reveals a novel AL2 silencing suppressor activity.

UNLABELLED: Both posttranscriptional and transcriptional gene silencing (PTGS and TGS, respectively) participate in defense against the DNA-containing geminiviruses. As a countermeasure, members of the genus Begomovirus (e.g., Cabbage leaf curl virus) encode an AL2 protein that is both a transcriptional activator and a silencing suppressor. The related L2 protein of Beet curly top virus (genus Curtovirus) lacks transcription activation activity. Previous studies showed that both AL2 and L2 suppress silencing by a mechanism that correlates with adenosine kinase (ADK) inhibition, while AL2 in addition activates transcription of cellular genes that negatively regulate silencing pathways. The goal of this study was to clarify the general means by which these viral proteins inhibit various aspects of silencing. We confirmed that AL2 inhibits systemic silencing spread by a mechanism that requires transcription activation activity. Surprisingly, we also found that reversal of PTGS and TGS by ADK inactivation depended on whether experiments were conducted in vegetative or reproductive Nicotiana benthamiana plants (i.e., before or after the vegetative-to-reproductive transition). While AL2 was able to reverse silencing in both vegetative and reproductive plants, L2 and ADK inhibition were effective only in vegetative plants. This suggests that silencing maintenance mechanisms can change during development or in response to stress. Remarkably, we also observed that AL2 lacking its transcription activation domain could reverse TGS in reproductive plants, revealing a third, previously unsuspected AL2 suppression mechanism that depends on neither ADK inactivation nor transcription activation. IMPORTANCE: RNA silencing in plants is a multivalent antiviral defense, and viruses respond by elaborating multiple and sometimes multifunctional proteins that inhibit various aspects of silencing. The studies described here add an additional layer of complexity to this interplay. By examining geminivirus AL2 and L2 suppressor activities, we show that L2 is unable to suppress silencing in Nicotiana benthamiana plants that have undergone the vegetative-to-reproductive transition. As L2 was previously shown to be effective in mature Arabidopsis plants, these results illustrate that silencing mechanisms can change during development or in response to stress in ways that may be species specific. The AL2 and L2 proteins are known to share a suppression mechanism that correlates with the ability of both proteins to inhibit ADK, while AL2 in addition can inhibit silencing by transcriptionally activating cellular genes. Here, we also provide evidence for a third AL2 suppression mechanism that depends on neither transcription activation nor ADK inactivation. In addition to revealing the remarkable versatility of AL2, this work highlights the utility of viral suppressors as probes for the analysis of silencing pathways.

Variation in management of lymph node positive prostate cancer after radical prostatectomy within a statewide quality improvement consortium.

BACKGROUND: Patients with lymph node positive (pN+) disease found at the time of radical prostatectomy with pelvic lymphadenectomy for clinically localized prostate cancer (CaP) are at high risk of disease persistence and progression. Contemporary management trends of pN+ CaP are not well described. MATERIALS AND METHODS: Patients in the Michigan Urologic Surgery Improvement Collaborative (MUSIC) with clinically localized prostate cancer who underwent radical prostatectomy between 2012 and 2023 with cN0/pN+ disease were identified. The primary outcome was to evaluate patient and practice-level factors associated with time to secondary post-RP treatment. Secondary outcomes included practice-level variation in management of pN+ CaP and rates of secondary treatment modality. To assess factors associated with secondary treatment, a Cox proportional hazards model of a 60-day landmark analysis was performed. RESULTS: We identified 666 patients with pN+ disease. Overall, 66% underwent secondary treatment within 12 months post-RP. About 19% of patients with detectable post-RP PSA did not receive treatment. Of patients receiving secondary treatment after 60-days post-RP, 34% received androgen deprivation therapy (ADT) alone, 27% received radiation (RT) alone, 36% received combination, and 4% received other systemic therapies. In the multivariable model, pathologic grade group (GG)3 (HR 1.5; 95%CI: 1.05-2.14), GG4-5 (HR 1.65; 95%CI: 1.16-2.34), positive margins (HR 1.46; 95%CI: 1.13-1.88), and detectable postoperative PSA ≥0.1 ng/ml (HR 3.46; 95%CI: 2.61-4.59) were significantly associated with secondary post-RP treatment. There was wide variation in adjusted practice-level 12-month secondary treatment utilization (28%-79%). CONCLUSIONS: The majority pN+ patients receive treatment within 12 months post-RP which was associated with high-risk pathological features and post-RP PSA. Variation in management of pN+ disease highlights the uncertainty regarding the optimal management. Understanding which patients will benefit from secondary treatment, and which type, will be critical to minimize variation in care.

Assessing the Clinical Utility of Published Prostate Cancer Polygenic Risk Scores in a Large Biobank Data Set.

BACKGROUND AND OBJECTIVE: Polygenic risk scores (PRSs) have been developed to identify men with the highest risk of prostate cancer. Our aim was to compare the performance of 16 PRSs in identifying men at risk of developing prostate cancer and then to evaluate the performance of the top-performing PRSs in differentiating individuals at risk of aggressive prostate cancer. METHODS: For this case-control study we downloaded 16 published PRSs from the Polygenic Score Catalog on May 28, 2021 and applied them to Michigan Genomics Initiative (MGI) patients. Cases were matched to the Michigan Urological Surgery Improvement Collaborative (MUSIC) registry to obtain granular clinical and pathological data. MGI prospectively enrolls patients undergoing surgery at the University of Michigan, and MUSIC is a multi-institutional registry that prospectively tracks demographic, treatment, and clinical variables. The predictive performance of each PRS was evaluated using the area under the covariate-adjusted receiver operating characteristic curve (aAUC), and the association between PRS and disease aggressiveness according to prostate biopsy data was measured using logistic regression. KEY FINDINGS AND LIMITATIONS: We included 18 050 patients in the analysis, of whom 15 310 were control subjects and 2740 were prostate cancer cases. The median age was 66.1 yr (interquartile range 59.9-71.6) for cases and 56.6 yr (interquartile range 42.6-66.7) for control subjects. The PRS performance in predicting the risk of developing prostate cancer according to aAUC ranged from 0.51 (95% confidence interval 0.51-0.53) to 0.67 (95% confidence interval 0.66-0.68). By contrast, there was no association between PRS and disease aggressiveness. CONCLUSIONS AND CLINICAL IMPLICATIONS: Prostate cancer PRSs have modest real-world performance in identifying patients at higher risk of developing prostate cancer; however, they are limited in distinguishing patients with indolent versus aggressive disease. PATIENT SUMMARY: Risk scores using data for multiple genes (called polygenic risk scores) can identify men at higher risk of developing prostate cancer. However, these scores need to be refined to be able to identify men with the highest risk for clinically significant prostate cancer.

Impact of Gene Expression Classifier Testing on Adjuvant Treatment Following Radical Prostatectomy: The G-MINOR Prospective Randomized Cluster-crossover Trial.

BACKGROUND AND OBJECTIVE: Decipher is a tissue-based genomic classifier (GC) developed and validated in the post-radical prostatectomy (RP) setting as a predictor of metastasis. We conducted a prospective randomized controlled cluster-crossover trial assessing the use of Decipher to determine its impact on adjuvant treatment after RP. METHODS: Eligible patients had undergone RP within 9 mo of enrollment, had pT3-4 disease and/or positive surgical margins, and prostate-specific antigen <0.1 ng/ml. Centers were randomized to a sequence of 3-mo periods of either GC-informed care or usual care (UC). Cancer of the Prostate Risk Assessment Postsurgical (CAPRA-S) recurrence risk scores were provided to treating physicians and patients in all periods. KEY FINDINGS AND LIMITATIONS: Impact of GC test results on adjuvant treatment were compared with UC alone. Longitudinal patient-reported urinary and sexual function was assessed. A total of 175 patients were enrolled in 27 periods with GC and 163 in 28 periods with UC. At 18 mo after RP, an average patient in the GC arm received adjuvant treatment 9.7% of the time compared with 8.7% for an average individual in the UC arm (0.99% mean difference, 95% confidence interval [CI] -7.6%, 9.6%, p = 0.8). While controlling for CAPRA-S score, higher GC scores tended to result in an increased likelihood of adjuvant treatment that was not statistically significant (odds ratio [OR] = 1.35 per 0.1 increase in GC score, 95% CI 0.98-1.85, p = 0.066). Using the GC risk groups, reflecting clinical use, a high GC risk was associated with significantly higher odds of receiving adjuvant treatment (OR = 6.9, 95% CI 1.8, 26, p = 0.005) compared with a low GC score, adjusted for CAPRA-S score. There were no differences in patient-reported urinary and sexual function between the study arms. As oncologic outcomes are immature, the present data cannot address whether GC testing provides any cancer control benefit. CONCLUSIONS AND CLINICAL IMPLICATIONS: GC testing impacts adjuvant therapy administration when viewed through the risk categories presented in the patient report; however, these data do not provide specific support for GC testing in the adjuvant treatment setting.

Integrative multi-region molecular profiling of primary prostate cancer in men with synchronous lymph node metastasis.

Localized prostate cancer is frequently composed of multiple spatially distinct tumors with significant inter- and intra-tumoral molecular heterogeneity. This genomic diversity gives rise to many competing clones that may drive the biological trajectory of the disease. Previous large-scale sequencing efforts have focused on the evolutionary process in metastatic prostate cancer, revealing a potential clonal progression to castration resistance. However, the clonal origin of synchronous lymph node (LN) metastases in primary disease is still unknown. Here, we perform multi-region, targeted next generation sequencing and construct phylogenetic trees in men with prostate cancer with synchronous LN metastasis to better define the pathologic and molecular features of primary disease most likely to spread to the LNs. Collectively, we demonstrate that a combination of histopathologic and molecular factors, including tumor grade, presence of extra-prostatic extension, cellular morphology, and oncogenic genomic alterations are associated with synchronous LN metastasis.

Trials in the Key of G: Building Level 1 Evidence on the Real-world Effectiveness of Prostate Biomarkers.

A number of genomic classifiers are available to aid in shared decision-making for men with localized prostate cancer; however, there is no high-level evidence assessing their clinical utility. The two randomized controlled trials in this report prospectively evaluate the use of gene expression classifier testing at the time of cancer diagnosis and after surgical treatment.

Prostate Cancer With Peritoneal Carcinomatosis: A Robotic-assisted Radical Prostatectomy-based Case Series.

OBJECTIVE: To aid in the diagnosis and treatment of patients with metastatic tumor seeding, an exceedingly rare phenomenon following minimally invasive urological surgery, additional case reports are needed. MATERIALS AND METHODS: We report our experience with patients determined to have peritoneal carcinomatosis following robotic-assisted radical prostatectomy (RARP) and provide a descriptive summary of these unique cases. RESULTS: Five cases of peritoneal carcinomatosis were identified, all of which occurred relatively late-between 8 and 13 years-following RARP. Four of the 5 cases had T3 disease at the time of prostatectomy. 68Ga-PSMA PET identified peritoneal carcinomatosis in 3 of 5 cases. CONCLUSION: Certain clinical factors, such as advanced pathologic stage at the time of prostatectomy, may predict risk for carcinomatosis following RARP. Additionally, next-generation imaging modalities, such as PSMA PET, may aid in identifying these metastases and are likely to identify increasing numbers of these patients as next-generation imaging becomes more widely available. Continued documentation and classification of this atypical presentation are needed to improve our understanding and management of this phenomenon.

Implementation of patient-reported outcome measures into health care for men with localized prostate cancer.

Measuring treatment-related quality of life (QOL) has become an increasingly requisite component of delivering high-quality care for patients with prostate cancer. Patient-reported outcome measures (PROMs) have, therefore, become an important tool for understanding the adverse effects of radical prostate cancer treatment and have been widely integrated into clinical practice. By providing real-time symptom monitoring and improved clinical feedback to patients and providers, PRO assessment has led to meaningful gains in prostate cancer care delivery and quality improvement worldwide. By providing an avenue for benchmarking, collaboration and population health monitoring, PROMs have delivered substantial improvements beyond providing individual symptom feedback. However, multilevel barriers exist that need to be addressed before the routine implementation of PROMs is achieved. Improvements in collection, interpretation, standardization and reporting will be crucial for the continued implementation of PROM instruments in prostate cancer pathways.

Urinary MyProstateScore (MPS) to Rule out Clinically-Significant Cancer in Men with Equivocal (PI-RADS 3) Multiparametric MRI: Addressing an Unmet Clinical Need.

OBJECTIVE: To evaluate the complementary value of urinary MyProstateScore (MPS) testing and multiparametric MRI (mpMRI) and assess outcomes in patients with equivocal mpMRI. MATERIALS AND METHODS: Included patients underwent mpMRI followed by urine collection and prostate biopsy at the University of Michigan between 2015 -2019. MPS values were calculated from urine specimens using the validated model based on serum PSA, urinary PCA3, and urinary TMPRSS2:ERG. In the PI-RADS 3 population, the discriminative accuracy of PSA, PSAD, and MPS for GG≥2 cancer was quantified by the AUC curve. Decision curve analysis was used to assess net benefit of MPS relative to PSAD. RESULTS: There were 540 patients that underwent mpMRI and biopsy with MPS available. The prevalence of GG≥2 cancer was 13% for PI-RADS 3, 56% for PI-RADS 4, and 87% for PI-RADS 5. MPS was significantly higher in men with GG≥2 cancer [median 44.9, IQR (29.4 -57.5)] than those with negative or GG1 biopsy [median 29.2, IQR (14.8 -44.2); P <.001] in the overall population and when stratified by PI-RADS score. In the PI-RADS 3 population (n = 121), the AUC for predicting GG≥2 cancer was 0.55 for PSA, 0.62 for PSAD, and 0.73 for MPS. MPS provided the highest net clinical benefit across all pertinent threshold probabilities. CONCLUSION: In patients that underwent mpMRI and biopsy, MPS was significantly associated with GG≥2 cancer across all PI-RADS scores. In the PI-RADS 3 population, MPS significantly outperformed PSAD in ruling out GG≥2 cancer. These findings suggest a complementary role of MPS testing in patients that have undergone mpMRI.

Understanding the Barriers to Neoadjuvant Chemotherapy in Patients with Muscle Invasive Bladder Cancer: A Quality Improvement Initiative.

PURPOSE: Utilization of neoadjuvant chemotherapy (NAC) for the management of muscle-invasive bladder cancer remains low. We sought to understand our practice of NAC use in order to design a quality improvement initiative geared towards optimizing medical oncology referral. MATERIALS AND METHODS: We identified 339 patients with ≥cT2 bladder cancer treated with radical cystectomy between 2012-2017 at our institution. We assessed the rate of referral to medical oncology, rate of NAC administration, as well as medical, patient and provider variables associated with NAC use. Bayesian logistic regression modeling identified variables associated with NAC use and chart review provided granular patient-level data. RESULTS: 85% (n=289) of patients were referred to medical oncology and 62.5% (n=212) received NAC. Renal insufficiency, hearing loss, and treating urologist were conclusively associated with lower odds of NAC use. 46 patients were not referred to medical oncology and 50% of these had medical contraindications to cisplatin cited as the reason for no referral. 38 patients met with medical oncology but did not receive NAC. 30 (79%) had comorbidities that impacted this decision with 15 (39%) ineligible based on impaired renal function. CONCLUSIONS: Despite the relatively high rates of medical oncology referral and NAC use in this cohort, there are still opportunities to improve the efficiency of this practice. Quality improvement initiatives could optimize the referral of patients with ≥T2 bladder cancer for consideration of cisplatin-based NAC and establish an important quality metric in the management of these patients.

Rethinking the one-size-fits-most approach to venous thromboembolism prophylaxis after radical cystectomy.

OBJECTIVES: Empirically dosed enoxaparin is routinely given in the postoperative period for venous thromboembolism (VTE) prophylaxis after radical cystectomy (RC). Patient-specific factors may alter its pharmacokinetics, and it is unclear whether this leads to levels sufficient for antithrombosis. We sought to evaluate variability of anti-factor Xa levels in a cohort of RC patients receiving perioperative enoxaparin prophylaxis. MATERIAL AND METHODS: Patients undergoing RC at a single institution were placed on a postoperative pathway that included enoxaparin. An anti-factor Xa level was drawn 2 to 4 hours after the third dose. The target range for prophylaxis was 0.3 IU/ml to 0.5 IU/ml. RESULTS: The primary outcome was anti-factor Xa level. Demographics, operative time, hospital course, and 30-days post-operative VTE were compared by anti-factor Xa level group using univariate and multivariable analyses. Between January 2018 and 2019, 107 RC patients remained on pathway and were included in our analysis. Sixty-five (61%) were below target range for VTE prophylaxis.  A single VTE event (0.9%) occurred in a subprophylactic individual. The subprophylactic group had a significantly higher body mass index (P < 0.01) than those within target range. CONCLUSIONS: Higher body mass index was associated with subprophylactic enoxaparin dosing after RC. Nearly two-thirds of patients had below target anti-factor Xa levels. This suggests that dosing could be further individualized, but given the low incidence of VTE, implications of dose-adjusted prophylaxis on VTE prevention remain unknown.

Overtreatment and Underutilization of Watchful Waiting in Men With Limited Life Expectancy: An Analysis of the Michigan Urological Surgery Improvement Collaborative Registry.

OBJECTIVE: To determine rates of watchful waiting (WW) vs treatment in prostate cancer (PCa) and limited life expectancy (LE) and assess determinants of management. MATERIALS AND METHODS: Patients diagnosed with PCa between 2012 and 2018 with <10 years LE were identified from the Michigan Urologic Surgery Improvement Collaborative registry. Multinomial logistic regression models were used to identify factors associated with management choice among NCCN low-risk PCa patients. Data from high-volume practices were analyzed to understand practice variation. RESULTS: Total 2393 patients were included. Overall, WW was performed in 8.1% compared to 23.3%, 25%, 11.2%, and 3.6% who underwent AS, radiation (XRT), prostatectomy (RP), and brachytherapy (BT), respectively. In men with NCCN low-risk disease (n = 358), WW was performed in 15.1%, compared to AS (69.3%), XRT (4.2%), RP (6.7%), and BT (2.5%). There was wide variation in management among practices in low-risk men; WW (6%-35%), AS (44%-81%), and definitive treatment (0%-30%). Older age was associated with less likelihood of undergoing AS vs WW (odds ratio [OR] 0.88, P < .001) or treatment vs WW (OR 0.83, P < .0001). Presence of ≥cT2 disease (OR 8.55, P = .014) and greater number of positive biopsy cores (OR 1.41, P = .014) was associated with greater likelihood of treatment vs WW and Charlson comorbidity score of 1 vs 0 (OR 0.23, P = .043) was associated with less likelihood of treatment vs WW. CONCLUSION: Wide practice level variation exists in management for patients with low- and favorable-risk PCa and <10-year LE. Utilization of WW is poor, suggesting overtreatment in men who will experience little benefit.

Wnt Signaling Drives Prostate Cancer Bone Metastatic Tropism and Invasion.

Wnt signaling has been implicated as a driver of prostate cancer-related osteoblast differentiation, and previous studies have linked modifications in Wnt function with the induction of tumor metastasis. A unique aspect of prostate cancer bone metastases in mouse models is their relative predilection to the hindlimb (femur) compared to the forelimb (humerus). Comparative gene expression profiling was performed within the humerus and femur from non-tumor-bearing mice to evaluate differences in the microenvironments of these locations. This revealed the relative overexpression of the Wnt signaling inhibitors WIF1 and SOST in the humerus compared to the femur, with increased WNT5A expression in femur bone marrow, suggesting a coordinated upregulation of Wnt signals within the femur compared to the humerus. Conditioned medium (CM) from bone marrow stromal cells (HS-5 cells) was used to mimic the bone marrow microenvironment, which strongly promoted prostate cancer cell invasion (3.3-fold increase in PC3 cells, P < .05; 7-fold increase in LNCaP cells, P < .05). WNT5A shRNA knockdown within the CM-producing HS-5 cells significantly decreased PC3 (56%, P < .05) and LNCaP (60%, P < .05) cell invasion. Similarly, preincubation of CM with WIF1 significantly blocked LNCaP cell invasion (40%, P < .05). shRNA-mediated knockdown of the Wnt receptors FZD4 and FZD8 also strongly inhibited tumor cell invasion (60% inhibition shFZD4, P < .05; 63% shFZD8, P < .05). Furthermore, small molecule inhibition of JNK, which is an important component of the noncanonical Wnt signaling pathway, significantly inhibited CM-mediated tumor invasion. Overall, this study reveals a role for Wnt signaling as a driver of prostate cancer bone metastatic tropism and invasion.

Individual and Population Comparisons of Surgery and Radiotherapy Outcomes in Prostate Cancer Using Bayesian Multistate Models.

Importance: Whether surgery or radiotherapy is the preferred treatment for patients with localized prostate cancer continues to be debated, and randomized clinical trials cannot yet fully address this question. Furthermore, there may be heterogeneity in responses, and the optimal treatment for a patient will depend on his clinical and tumor characteristics. Objectives: To use a unified statistical approach to compare the association of surgery and radiotherapy with both metastatic clinical failure (CF) and survival in localized prostate cancer and to develop an online calculator for individualized, treatment-specific outcome prediction. Design, Setting, and Participants: Cohort study for statistical analysis and development of individualized predictions using Bayesian multistate models that jointly consider both CF and survival and adjust for confounding factors. This study used data from patients treated at the University of Michigan between January 1, 1996, and July 1, 2013, with detailed information on treatment, patient and tumor characteristics, and outcomes. Primary analyses were performed in 2017 and 2018. Participants were a cohort of 4544 patients with localized prostate cancer undergoing primary treatment. Exposures: Radical prostatectomy and external beam radiotherapy. Main Outcomes and Measures: The clinical outcomes were metastatic CF, death after CF, and death from other causes. The adjustment factors were age, prostate gland volume, prostate-specific antigen level, comorbidities, Gleason score, perineural invasion, cT category, race, and treatment year. An online calculator was developed to estimate risks for multiple outcomes for any patient based on 2 treatment choices and on his clinical and tumor characteristics. Results: Among 4544 men (mean [SD] age, 61.2 [8.0] years), 3769 underwent radical prostatectomy, 775 received external beam radiotherapy, 157 (3.5%) had CF, 90 (2.0%) died after CF, and 378 (8.3%) died of other causes. Across all patients, there was no significant difference in risk of CF for surgery vs radiotherapy (hazard ratio, 0.80; 95% CI, 0.52-1.23). However, using multistate models, in some cases individualized predictions resulted in different expected outcomes between surgery and radiotherapy for a given patient. Conclusions and Relevance: In this study, after adjustment for measured confounders, the hazard of CF was similar between treatments on average. However, these data indicate a greater oncologic benefit for some individual patients if treated with surgery and for other patients if treated with radiotherapy. Individualized predictions provide a novel approach to facilitate treatment decision making.