| 1. |
- Spjuth, Ola, 1977-, et al.
(författare)
-
E-Science technologies in a workflow for personalized medicine using cancer screening as a case study
- 2017
-
Ingår i: JAMIA Journal of the American Medical Informatics Association. - : Oxford University Press. - 1067-5027 .- 1527-974X. ; 24:5, s. 950-957
-
Tidskriftsartikel (refereegranskat)abstract
- Objective: We provide an e-Science perspective on the workflow from risk factor discovery and classification of disease to evaluation of personalized intervention programs. As case studies, we use personalized prostate and breast cancer screenings.Materials and Methods: We describe an e-Science initiative in Sweden, e-Science for Cancer Prevention and Control (eCPC), which supports biomarker discovery and offers decision support for personalized intervention strategies. The generic eCPC contribution is a workflow with 4 nodes applied iteratively, and the concept of e-Science signifies systematic use of tools from the mathematical, statistical, data, and computer sciences.Results: The eCPC workflow is illustrated through 2 case studies. For prostate cancer, an in-house personalized screening tool, the Stockholm-3 model (S3M), is presented as an alternative to prostate-specific antigen testing alone. S3M is evaluated in a trial setting and plans for rollout in the population are discussed. For breast cancer, new biomarkers based on breast density and molecular profiles are developed and the US multicenter Women Informed to Screen Depending on Measures (WISDOM) trial is referred to for evaluation. While current eCPC data management uses a traditional data warehouse model, we discuss eCPC-developed features of a coherent data integration platform.Discussion and Conclusion: E-Science tools are a key part of an evidence-based process for personalized medicine. This paper provides a structured workflow from data and models to evaluation of new personalized intervention strategies. The importance of multidisciplinary collaboration is emphasized. Importantly, the generic concepts of the suggested eCPC workflow are transferrable to other disease domains, although each disease will require tailored solutions.
|
|
| 2. |
|
|
| 3. |
- Pálsdóttir, Þorgerður
(författare)
-
Risk prediction in prostate cancer diagnostics : current challenges and improvements
- 2019
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Prostate cancer has been considered a disease of elderly men, and thus historically less focus has been on prostate cancer research than many other cancer types. However, as life expectancy is increasing all over the world, more life years are lost when men are diagnosed with prostate cancer at the age of 70 years now than before. Therefore, it is increasingly important to improve the diagnostic pathway of prostate cancer in modern health care. My thesis aims to address some of the issues in the current prostate cancer diagnostic pipeline using risk prediction mod-els. Measuring the level of prostate-specific antigen (PSA) in blood is widely used as a blood test to screen for prostate cancer and evidence has shown that mortality decreases with PSA testing. However, because PSA testing has a high false-positive rate, many unnecessary biopsies are per-formed on healthy men and many men are overdiagnosed with indolent disease (International Society of Urological Pathology (ISUP) grade group 1). In Study I the objective was to predict the risk of clinically consequential cancer (ISUP ≥ 2) at biopsy and the cumulative probability of having a negative biopsy when being PSA tested with one, two, three, four, or five to eight year intervals. We found that men with a PSA level above 1 ng/mL had an increased risk of ISUP ≥ 2 prostate cancer when screened with longer then annual intervals, while men with a PSA level below 1 ng/mL had low risk of ISUP ≥ 2 prostate cancer regardless of time between testing. The benefit of a shorter screening interval needs to be balanced with the increased cumulative probability of having a negative biopsy which we found to be twofold for annual vs. biennial testing intervals and threefold for annual vs. triennial testing intervals. Knowledge about the relationship between PSA, age and different grades of prostate cancer is important for clinicians working with prostate cancer diagnosis because of how widely used the PSA test is. In Study II we studied the association between the risk of indolent and clinically consequential prostate cancer (ISUP 1 and ISUP ≥ 2) and PSA and age, respectively. Our study cohort comprised of 6.083 biopsied men from the STHLM3 study and 72.996 biopsy cores from those men. In the overall ISUP grade system, lower grades can be masked by higher grades, and thus we studied the associations for both overall ISUP grade and for ISUP grade on each biopsy core. Our results showed that ISUP 1 prostate cancer was not significantly associated with PSA or age, on overall ISUP grade or on individual biopsy core level. In contrast, our results showed that ISUP ≥ 2 prostate cancer is significantly associated with increasing PSA level and older age. Our results indicate that PSA leakage of ISUP 1 prostate cancer cells is more similar to that of benign prostate tissue than ISUP ≥ 2 prostate cancer tissue. The use of magnetic resonance imaging (MRI) before biopsy to diagnose prostate cancer has increased in current clinical practice. Combining results from prostate MRI with existing risk prediction models can improve the predictive abilities of the models. The aim of Study III was to develop a risk prediction model (S3M-MRI), combining the Stockholm3 score and the PI-RADS (Prostate Imaging Reporting and data System) score from MRI to predict the risk of ISUP ≥ 2 prostate cancer. We developed the S3M-MRI model using data from the STHLM3-MRI diagnostic study and compared the model performance of the S3M-MRI to the Stockholm3 model and PI-RADS score. We also compared five diagnostic strategies for clinical outcomes. We found that the combined S3M-MRI model had better predictive abilities than both the Stockholm3 and the PI-RADS alone. However, when we compared it to different clinical strate-gies, the sequential use of the Stockholm3 test followed by MRI on Stockholm3 positive men resulted in similar numbers of performed biopsies and diagnosed ISUP ≥ 2 prostate cancers while saving many MRI scans. Prostate cancer diagnosis is based on the result of the prostate biopsy and reclassification of ISUP grade on radical prostatectomy samples compared to biopsy is common. In Study IV our aim was to study what effect reclassification of disease status based on prostate biopsies has on the performance of prostate cancer risk prediction models using simulations and data from the STHLM3 Radical Prostatectomy Cohort. The cohort comprised of 780 men from the STHLM3 study who were diagnosed with prostate cancer and treated with a radical prostatec-tomy between 2013 and 2015. We compared four simulated prediction model scenarios with and without error in disease status and calculated the area under the receiver operating charac-teristics (ROC) curve (AUC) of the Stockholm3 score for predicting clinically significant prostate cancer assessed using biopsy and radical prostatectomy samples. Our simulations showed that fitting a risk prediction model using data with error in the disease status only leads to a small decline in the true predictive performance, but leads to a large decline in apparent predictive performance when evaluated against data with error in the disease status. Moreover, our results showed that the Stockholm3 test has stronger association with clinically significant prostate cancer defined on prostatectomy samples (without errors) than on biopsy samples (with errors). In conclusion, in this thesis we have aimed to describe a part of the risk associated with diag-nosis of prostate cancer as well as developing new prostate cancer risk prediction models. This thesis contributes to the constant pursue of improving the current prostate cancer diagnostic pipeline in order to improve the lives of men screened for or diagnosed with prostate cancer.
|
|
| 4. |
- Waldén, Mauritz, et al.
(författare)
-
A Head-to-head Comparison of Prostate Cancer Diagnostic Strategies Using the Stockholm3 Test, Magnetic Resonance Imaging, and Swedish National Guidelines : Results from a Prospective Population-based Screening Study
- 2022
-
Ingår i: European Urology Open Science. - : Elsevier. - 2666-1691 .- 2666-1683. ; 38, s. 32-39
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Strategies for early detection of prostate cancer aim to detect clinically significant prostate cancer (csPCa) and avoid detection of insignificant cancers and unnecessary biopsies. Swedish national guidelines (SNGs), years 2019 and 2020, involve prostate-specific antigen (PSA) testing, clinical variables, and magnetic resonance imaging (MRI). The Stockholm3 test and MRI have been suggested to improve selection of men for prostate biopsy. Performance of SNGs compared with the Stockholm3 test or MRI in a screening setting is unclear.Objective: To compare strategies based on previous and current national guidelines, Stockholm3, and MRI to select patients for biopsy in a screening-by-invitation setting.Design setting and participants: All participants underwent PSA test, and men with PSA ≥3 ng/ml underwent Stockholm3 testing and MRI. Men with Stockholm3 ≥11%, Prostate Imaging Reporting and Data System score ≥3 on MRI, or indication according to SNG-2019 or SNG-2020 were referred to biopsy.Outcome measurements and statistical analysis: The primary outcome was the detection of csPCa at prostate biopsy, defined as an International Society of Urological Pathology (ISUP) grade of ≥2.Results and limitations: We invited 8764 men from the general population, 272 of whom had PSA ≥3 ng/ml. The median PSA was 4.1 (interquartile range: 3.4-5.8), and 136 of 270 (50%) who underwent MRI lacked any pathological lesions. In total, 37 csPCa cases were diagnosed. Using SNG-2019, 36 csPCa cases with a high biopsy rate (179 of 272) were detected and 49 were diagnosed with ISUP 1 cancers. The Stockholm3 strategy diagnosed 32 csPCa cases, with 89 biopsied and 27 ISUP 1 cancers. SNG-2020 detected 32 csPCa and 33 ISUP 1 cancer patients, with 99 men biopsied, and the MRI strategy detected 30 csPCa and 35 ISUP 1 cancer cases by biopsying 123 men. The latter two strategies generated more MRI scans than the Stockholm3 strategy (n = 270 vs 33).Conclusions: Previous guidelines provide high detection of significant cancer but at high biopsy rates and detection of insignificant cancer. The Stockholm3 test may improve diagnostic precision compared with the current guidelines or using only MRI.Patient summary: The Stockholm3 test facilitates detection of significant cancer, and reduces the number of biopsies and detection of insignificant cancer.
|
|
