Timothy J. Daskivich

Questioning the 10-year Life Expectancy Rule for High-grade Prostate Cancer: Comparative Effectiveness of Aggressive vs Nonaggressive Treatment of High-grade Disease in Older Men With Differing Comorbid Disease Burdens

OBJECTIVE To determine if the 10-year rule should apply to men with high-grade, clincially localized prostate cancer, we characterized the survival benefits of aggressive (surgery, radiation, brachytherapy) over nonaggressive treatment (watchful waiting, active surveillance) among older men with differing comorbidity at diagnosis. METHODS We sampled 44,521 men older than 65 with cT1-2, poorly differentiated prostate cancer diagnosed in 1991-2007 from the Surveillance, Epidemiology, and End Results-Medicare database. We used propensity-adjusted, competing-risks regression to calculate 5- and 10-year cancer mortality among those treated aggressively and nonaggressively across comorbidity subgroups. We determined 5- and 10-year absolute risk reduction in cancer mortality and numbers needed to treat to prevent one cancer death at 10 years. RESULTS In propensity-adjusted, competing-risks regression analysis, aggressive treatment was associated with significantly lower risk of cancer mortality for those with Charlson scores of 0 (sub-hazard ratio (SHR) 0.43, 95% confidence interval [CI] 0.39-0.47), 1 (SHR 0.48, 95% CI 0.40-0.58), and 2 (SHR 0.46, 95% CI 0.34-0.62) but not 3+ (SHR 0.68, 95% CI 0.44-1.07). Absolute reductions in cancer mortality between those treated aggressively and nonaggressively were 7%, 5.5%, 6.9%, and 2.5% at 5 years, and 11.3%, 7.9%, 8.6%, and 2.8% at 10 years for men with Charlson scores of 0, 1, 2, and 3+ , respectively; numbers needed to treat to prevent 1 cancer death at 10 years were 9, 13, 12, and 36 men. CONCLUSION The 10-year rule may not apply to men with high-grade, clinically localized disease. Older men with Charlson scores ≤2 should consider aggressive treatment of such disease due to its substantial short-term cancer survival benefits.

Online physician ratings fail to predict actual performance on measures of quality, value, and peer review

Objective Patients use online consumer ratings to identify high-performing physicians, but it is unclear if ratings are valid measures of clinical performance. We sought to determine whether online ratings of specialist physicians from 5 platforms predict quality of care, value of care, and peer-assessed physician performance. Materials and Methods We conducted an observational study of 78 physicians representing 8 medical and surgical specialties. We assessed the association of consumer ratings with specialty-specific performance scores (metrics including adherence to Choosing Wisely measures, 30-day readmissions, length of stay, and adjusted cost of care), primary care physician peer-review scores, and administrator peer-review scores. Results Across ratings platforms, multivariable models showed no significant association between mean consumer ratings and specialty-specific performance scores (β-coefficient range, -0.04, 0.04), primary care physician scores (β-coefficient range, -0.01, 0.3), and administrator scores (β-coefficient range, -0.2, 0.1). There was no association between ratings and score subdomains addressing quality or value-based care. Among physicians in the lowest quartile of specialty-specific performance scores, only 5%-32% had consumer ratings in the lowest quartile across platforms. Ratings were consistent across platforms; a physicians score on one platform significantly predicted his/her score on another in 5 of 10 comparisons. Discussion Online ratings of specialist physicians do not predict objective measures of quality of care or peer assessment of clinical performance. Scores are consistent across platforms, suggesting that they jointly measure a latent construct that is unrelated to performance. Conclusion Online consumer ratings should not be used in isolation to select physicians, given their poor association with clinical performance.

Life Expectancy and Variation in Treatment for Early Stage Kidney Cancer

PURPOSE Patients with limited life expectancy are at risk for overtreatment of T1a kidney cancer. We sought to determine patterns of treatment for T1a kidney cancer in a nationally representative sample of patients with life expectancy less than 10 and less than 5 years. MATERIALS AND METHODS We sampled 9,825 patients older than 65 years with clinical T1a kidney cancer diagnosed between 2000 and 2010 from the SEER (Surveillance, Epidemiology and End Results)-Medicare database. We performed competing risks regression to model survival by age/comorbidity and identified patients with life expectancy less than 10 and less than 5 years. Multivariate logistic regression was used to determine the probability of aggressive treatment with surgery or ablation among those with limited life expectancy. RESULTS Life expectancy was less than 10 years in patients 66 to 80 years old with a Charlson score of 3+, in those 80 to 84 years old with a Charlson score of 1+ and in all patients 85 years old or older. Among those with life expectancy less than 10 years the multivariate probability of aggressive treatment was 85%, 84%, 82%, 75% and 50% in those 66 to 69, 70 to 74, 75 to 79, 80 to 84 and 85 years old or older, respectively. In those with life expectancy less than 10 years who were treated aggressively treatment was radical nephrectomy in 61%, partial nephrectomy in 24% and ablation in 14%. Among those with life expectancy less than 5 years (age 85 years or greater with a Charlson score of 3+) the multivariate probability of aggressive treatment was 41% and more often surgery than ablation (68% vs 32% of patients). CONCLUSIONS The majority of patients with life expectancy less than 10 years and a significant minority with life expectancy less than 5 years were treated with surgery or ablation for T1a kidney cancer. Life expectancy should be better incorporated into treatment decision making for early stage kidney cancer.

Cancer and All-cause Mortality in Bladder Cancer Patients Undergoing Radical Cystectomy: Development and Validation of a Nomogram for Treatment Decision-making

OBJECTIVE To develop and validate a nomogram assessing cancer and all-cause mortality following radical cystectomy. Given concerns regarding the morbidity associated with surgery, there is a need for incorporation of cancer-specific and competing risks into patient counseling and recommendations. MATERIALS AND METHODS A total of 5325 and 1257 diagnosed with clinical stage T2-T4a muscle-invasive bladder cancer from January 1, 2006 to December 31, 2011 from Surveillance, Epidemiology, and End Results-Medicare and Texas Cancer Registry-Medicare linked data, respectively. Cox proportional hazards models were used and a nomogram was developed to predict 3- and 5-year overall and cancer-specific survival with external validation. RESULTS Patients who underwent radical cystectomy were mostly younger, male, married, non-Hispanic white and had fewer comorbidities than those who did not undergo radical cystectomy (P < .001). Married patients, in comparison with their unmarried counterparts, had both improved overall (hazard ratio 0.76; 95% confidence interval 0.70-0.83, P < .001) and cancer-specific (hazard ratio 0.76; 95% confidence interval 0.68-0.85, P < .001) survival. A nomogram developed using Surveillance, Epidemiology, and End Results-Medicare data, predicted 3- and 5-year overall and cancer-specific survival rates with concordance indices of 0.65 and 0.66 in the validated Texas Cancer Registry-Medicare cohort, respectively. CONCLUSION Older, unmarried patients with increased comorbidities are less likely to undergo radical cystectomy. We developed and validated a generalizable instrument that has been converted into an online tool (Radical Cystectomy Survival Calculator), to provide a benefit-risk assessment for patients considering radical cystectomy.

Patient function, long‐term survival, and use of surgery in patients with kidney cancer

Beyond age and comorbidity, functionality can shape the long‐term survival potential of patients with cancer. Accordingly, herein the authors compared mortality and receipt of cancer‐directed surgery according to patient function among older adults with kidney cancer.

Development of a nationally-representative coordinated registry network for prostate ablation technologies

Purpose: The accumulation of data through a prospective, multicenter coordinated registry network is a practical way to gather real world evidence on the performance of novel prostate ablation technologies. Urological oncologists, targeted biopsy experts, industry representatives and representatives of the FDA (Food and Drug Administration) convened to discuss the role, feasibility and important data elements of a coordinated registry network to assess new and existing prostate ablation technologies. Materials and Methods: A multiround Delphi consensus approach was performed which included the opinion of 15 expert urologists, representatives of the FDA and leadership from high intensity focused ultrasound device manufacturers. Stakeholders provided input in 3 consecutive rounds with conference calls following each round to obtain consensus on remaining items. Participants agreed that these elements initially developed for high intensity focused ultrasound are compatible with other prostate ablation technologies. Coordinated registry network elements were reviewed and supplemented with data elements from the FDA common study metrics. Results: The working group reached consensus on capturing specific patient demographics, treatment details, oncologic outcomes, functional outcomes and complications. Validated health related quality of life questionnaires were selected to capture patient reported outcomes, including the IIEF‐5 (International Index of Erectile Function‐5), the I‐PSS (International Prostate Symptom Score), the EPIC‐26 (Expanded Prostate Cancer Index Composite‐26) and the MSHQ‐EjD (Male Sexual Health Questionnaire for Ejaculatory Dysfunction). Group consensus was to obtain followup multiparametric magnetic resonance imaging and prostate biopsy approximately 12 months after ablation with additional imaging or biopsy performed as clinically indicated. Conclusions: A national prostate ablation coordinated registry network brings forth vital practice pattern and outcomes data for this emerging treatment paradigm in the United States. Our multiple stakeholder consensus identifies critical elements to evaluate new and existing energy modalities and devices.

Online Ratings Systems for Physicians and Institutions: Limitations of the Current State of the Art

Consumers are increasingly using online ratings tools that compare surgeons and institutions to identify high-quality providers. However, concerns regarding their limitations-data quality, validity of statistical comparisons, and impact on access to care-should be considered before full-scale implementation.

Genes involved in prostate cancer progression determine MRI visibility

MRI is used to image prostate cancer and target tumors for biopsy or therapeutic ablation. The objective was to understand the biology of tumors not visible on MRI that may go undiagnosed and untreated. Methods: Prostate cancers visible or invisible on multiparametric MRI were macrodissected and examined by RNAseq. Differentially expressed genes (DEGs) based on MRI visibility status were cross-referenced with publicly available gene expression databases to identify genes associated with disease progression. Genes with potential roles in determining MRI visibility and disease progression were knocked down in murine prostate cancer xenografts, and imaged by MRI. Results: RNAseq identified 1,654 DEGs based on MRI visibility status. Comparison of DEGs based on MRI visibility and tumor characteristics revealed that Gleason score (dissimilarity test, p<0.0001) and tumor size (dissimilarity test, p<0.039) did not completely determine MRI visibility. Genes in previously reported prognostic signatures significantly correlated with MRI visibility suggesting that MRI visibility was prognostic. Cross-referencing DEGs with external datasets identified four genes (PHYHD1, CENPF, ALDH2, GDF15) that predict MRI visibility, progression free survival and metastatic deposits. Genetic modification of a human prostate cancer cell line to induce miR-101 and suppress CENPF decreased cell migration and invasion. As prostate cancer xenografts in mice, these cells had decreased visibility on diffusion weighted MRI and decreased perfusion, which correlated with immunostaining showing decreased cell density and proliferation. Conclusions: Genes involved in prostate cancer prognosis and metastasis determine MRI visibility, indicating that MRI visibility has prognostic significance. MRI visibility was associated with genetic features linked to poor prognosis.

Limitations of the National Comprehensive Cancer Network® (NCCN®) Guidelines for Prediction of Limited Life Expectancy in Men with Prostate Cancer

Purpose: National Comprehensive Cancer Network prostate cancer guidelines for the prediction of life expectancy recommend subtracting 50% of life table predicted longevity for those in the lowest quartile of health. However, it is unclear how to identify these men and if their survival is uniform. Materials and Methods: We sampled records of 1,482 men diagnosed with prostate cancer from 1998 to 2004 at 2 VA hospitals. We identified men in the lowest quartile of health by age using Charlson scores, calculated their NCCN predicted life expectancy, and compared this with observed median survival in aggregate and across comorbidity subgroups. Results: Men with Charlson scores of 2+ (age less than 75 years) and 3+ (age 75 years or older) comprised the lowest quartile of health. Among those younger than 65, 65 to 69, 70 to 74, 75 to 79 and 80 years or older, observed survival vs NCCN predicted life expectancy in years was similar at 10.4 vs 11.1, 10.0 vs 7.8, 6.2 vs 6.4, 4.4 vs 4.9 and 3.7 vs 3.3, respectively. Yet within the lowest quartile there was significant heterogeneity in survival among men with differing Charlson scores. For example, men age 65 to 69 years with Charlson scores 2, 3 and 4+ had an observed median survival greater than 13.3, 9.4 and 4.3 years, respectively. NCCN guidelines misclassified 10‐year life expectancy in 24% and 56% of men age less than 65 and 65 to 69 years, and 5‐year life expectancy in 18% of men age 70 to 74 years. Conclusions: While NCCN predictions matched observed survival on average for the lowest quartile of health, there was substantial heterogeneity in survival by Charlson scores. More granular assessments of life expectancy should be used for those at highest risk for mortality.

Association of comorbid disease burden at diagnosis with higher tumor grade in men with non-metastatic prostate cancer

Background:While older age is associated with higher tumor grade, it is unknown whether comorbid disease burden has a similar, independent association. We sought to evaluate the impact of comorbid disease burden on tumor grade at diagnosis as indicated by biopsy Gleason score.Methods:We conducted an observational cohort study of 1260 men newly diagnosed with non-metastatic prostate cancer from 1998 to 2004 at two Veterans Affairs Medical Centers. Multivariable ordinal and multinomial logistic regression were used to evaluate the association between Charlson Comorbidity Index score and biopsy Gleason score.Results:Men with Charlson scores of 2 (odds ratio (OR) 1.8, P<0.001) and 3+ (OR 1.8, P<0.001) had significantly greater odds of higher Gleason scores, compared with men with Charlson scores of 0. In a multinomial logistic regression model predicting Gleason 7 vs ⩾6, only men with Charlson scores of 2 (OR 1.6, P=0.01) had greater odds of having a Gleason 7 tumor, compared with those with Charlson scores of 0. In a multinomial logistic regression model predicting Gleason 8–10 vs ⩽6, those with Charlson scores of 1 (OR 1.6, P=0.047), 2 (OR 2.8, P=0.01) and 3+ (OR 2.9, P=0.001) had higher odds of having a Gleason 8–10 tumor.Conclusions:Moderate-to-heavy comorbid disease burden at diagnosis may be associated with high tumor grade, independent of age, and is a stronger predictor of Gleason 8–10 than Gleason 7 disease.