Martin Eklund

Prostate cancer screening in men aged 50–69 years (STHLM3): a prospective population-based diagnostic study

BACKGROUND The prostate-specific antigen (PSA) test is used to screen for prostate cancer but has a high false-positive rate that translates into unnecessary prostate biopsies and overdiagnosis of low-risk prostate cancers. We aimed to develop and validate a model to identify high-risk prostate cancer (with a Gleason score of at least 7) with better test characteristics than that provided by PSA screening alone. METHODS The Stockholm 3 (STHLM3) study is a prospective, population-based, paired, screen-positive, diagnostic study of men without prostate cancer aged 50-69 years randomly invited by date of birth from the Swedish Population Register kept by the Swedish Tax Agency. Men with prostate cancer at enrolment were excluded from the study. The predefined STHLM3 model (a combination of plasma protein biomarkers [PSA, free PSA, intact PSA, hK2, MSMB, MIC1], genetic polymorphisms [232 SNPs], and clinical variables [age, family, history, previous prostate biopsy, prostate exam]), and PSA concentration were both tested in all participants enrolled. The primary aim was to increase the specificity compared with PSA without decreasing the sensitivity to diagnose high-risk prostate cancer. The primary outcomes were number of detected high-risk cancers (sensitivity) and the number of performed prostate biopsies (specificity). The STHLM3 training cohort was used to train the STHLM3 model, which was prospectively tested in the STHLM3 validation cohort. Logistic regression was used to test for associations between biomarkers and clinical variables and prostate cancer with a Gleason score of at least 7. This study is registered with ISCRTN.com, number ISRCTN84445406. FINDINGS The STHLM3 model performed significantly better than PSA alone for detection of cancers with a Gleason score of at least 7 (p<0·0001), the area under the curve was 0·56 (95% CI 0·55-0·60) with PSA alone and 0·74 (95% CI 0·72-0·75) with the STHLM3 model. All variables used in the STHLM3 model were significantly associated with prostate cancers with a Gleason score of at least 7 (p<0·05) in a multiple logistic regression model. At the same level of sensitivity as the PSA test using a cutoff of ≥3 ng/mL to diagnose high risk prostate cancer, use of the STHLM3 model could reduce the number of biopsies by 32% (95% CI 24-39) and could avoid 44% (35-54) of benign biopsies. INTERPRETATION The STHLM3 model could reduce unnecessary biopsies without compromising the ability to diagnose prostate cancer with a Gleason score of at least 7, and could be a step towards personalised risk-based prostate cancer diagnostic programmes. FUNDING Stockholm County Council (Stockholms Läns Landsting).

Bioclipse: an open source workbench for chemo- and bioinformatics

BackgroundThere is a need for software applications that provide users with a complete and extensible toolkit for chemo- and bioinformatics accessible from a single workbench. Commercial packages are expensive and closed source, hence they do not allow end users to modify algorithms and add custom functionality. Existing open source projects are more focused on providing a framework for integrating existing, separately installed bioinformatics packages, rather than providing user-friendly interfaces. No open source chemoinformatics workbench has previously been published, and no sucessful attempts have been made to integrate chemo- and bioinformatics into a single framework.ResultsBioclipse is an advanced workbench for resources in chemo- and bioinformatics, such as molecules, proteins, sequences, spectra, and scripts. It provides 2D-editing, 3D-visualization, file format conversion, calculation of chemical properties, and much more; all fully integrated into a user-friendly desktop application. Editing supports standard functions such as cut and paste, drag and drop, and undo/redo. Bioclipse is written in Java and based on the Eclipse Rich Client Platform with a state-of-the-art plugin architecture. This gives Bioclipse an advantage over other systems as it can easily be extended with functionality in any desired direction.ConclusionBioclipse is a powerful workbench for bio- and chemoinformatics as well as an advanced integration platform. The rich functionality, intuitive user interface, and powerful plugin architecture make Bioclipse the most advanced and user-friendly open source workbench for chemo- and bioinformatics. Bioclipse is released under Eclipse Public License (EPL), an open source license which sets no constraints on external plugin licensing; it is totally open for both open source plugins as well as commercial ones. Bioclipse is freely available at http://www.bioclipse.net.

Polygenic risk score improves prostate cancer risk prediction: results from the Stockholm-1 cohort study.

BACKGROUND More than 1 million prostate biopsies are conducted yearly in the United States. The low specificity of prostate-specific antigen (PSA) results in diagnostic biopsies in men without prostate cancer (PCa). Additional information, such as genetic markers, could be used to avoid unnecessary biopsies. OBJECTIVE To determine whether single nucleotide polymorphisms (SNPs) associated with PCa can be used to determine whether biopsy of the prostate is necessary. DESIGN, SETTINGS, AND PARTICIPANTS The Stockholm-1 cohort (n = 5241) consisted of men who underwent a prostate biopsy during 2005 to 2007. PSA levels were retrieved from databases and family histories were obtained using a questionnaire. Thirty-five validated SNPs were analysed and converted into a genetic risk score that was implemented in a risk-prediction model. RESULTS AND LIMITATIONS When comparing the nongenetic model (based on age, PSA, free-to-total PSA, and family history) with the genetic model and using a fixed number of detected PCa cases, it was found that the genetic model required significantly fewer biopsies than the nongenetic model, with 480 biopsies (22.7%) avoided, at a cost of missing a PCa diagnosis in 3% of patients characterised as having an aggressive disease. However, the overall genetic model does not discriminate between aggressive and nonaggressive cases. CONCLUSION Although the genetic model reduced the number of biopsies more than the nongenetic model, the clinical significance of this finding requires further evaluation.

Comparison Between the Four-kallikrein Panel and Prostate Health Index for Predicting Prostate Cancer

BACKGROUND The four-kallikrein panel and the Prostate Health Index (PHI) have been shown to improve prediction of prostate cancer (PCa) compared with prostate-specific antigen (PSA). No comparison of the four-kallikrein panel and PHI has been presented. OBJECTIVE To compare the four-kallikrein panel and PHI for predicting PCa in an independent cohort. DESIGN, SETTING, AND PARTICIPANTS Participants were from a population-based cohort of PSA-tested men in Stockholm County. We included 531 men with PSA levels between 3 and 15 ng/ml undergoing first-time prostate biopsy during 2010-2012. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS Models were fitted to case status. We computed calibration curves, the area under the receiver-operating characteristics curve (AUC), decision curves, and percentage of saved biopsies. RESULTS AND LIMITATIONS The four-kallikrein panel showed AUCs of 69.0 when predicting any-grade PCa and 71.8 when predicting high-grade cancer (Gleason score ≥7). Similar values were found for PHI: 70.4 and 71.1, respectively. Both models had higher AUCs than a base model with PSA value and age (p<0.0001 for both); differences between models were not significant. Sensitivity analyses including men with any PSA level or a previous biopsy did not materially affect our findings. Using 10% predicted risk of high-grade PCa by the four-kallikrein panel or PHI of 39 as cut-off for biopsy saved 29% of performed biopsies at a cost of delayed diagnosis for 10% of the men with high-grade cancers. Both models showed limited net benefit in decision analysis. The main study limitation was lack of digital rectal examination data and biopsy decision being based on PSA information. CONCLUSIONS The four-kallikrein panel and PHI similarly improved discrimination when predicting PCa and high-grade PCa. Both are simple blood tests that can reduce the number of unnecessary biopsies compared with screening with total PSA, representing an important new option to reduce harm. PATIENT SUMMARY Prostate-specific antigen screening is controversial due to limitations of the test. We found that two blood tests, the Prostate Health Index and the four-kallikrein panel, performed similarly and could both aid in decision making among Swedish men undergoing a prostate biopsy.

Proteochemometric modeling of HIV protease susceptibility

BackgroundA major obstacle in treatment of HIV is the ability of the virus to mutate rapidly into drug-resistant variants. A method for predicting the susceptibility of mutated HIV strains to antiviral agents would provide substantial clinical benefit as well as facilitate the development of new candidate drugs. Therefore, we used proteochemometrics to model the susceptibility of HIV to protease inhibitors in current use, utilizing descriptions of the physico-chemical properties of mutated HIV proteases and 3D structural property descriptions for the protease inhibitors. The descriptions were correlated to the susceptibility data of 828 unique HIV protease variants for seven protease inhibitors in current use; the data set comprised 4792 protease-inhibitor combinations.ResultsThe model provided excellent predictability (R2 = 0.92, Q2 = 0.87) and identified general and specific features of drug resistance. The models predictive ability was verified by external prediction in which the susceptibilities to each one of the seven inhibitors were omitted from the data set, one inhibitor at a time, and the data for the six remaining compounds were used to create new models. This analysis showed that the over all predictive ability for the omitted inhibitors was Q2inhibitors= 0.72.ConclusionOur results show that a proteochemometric approach can provide generalized susceptibility predictions for new inhibitors. Our proteochemometric model can directly analyze inhibitor-protease interactions and facilitate treatment selection based on viral genotype. The model is available for public use, and is located at HIV Drug Research Centre.

Aggregate Cost of Mammography Screening in the United States: Comparison of Current Practice and Advocated Guidelines

Context The total cost of alternative programs that screen for breast cancer is unknown. Contribution The actual cost of breast cancer screening in the United States in 2010 was

Bioclipse 2: A scriptable integration platform for the life sciences

7.8 billion. The cost would have been

Introducing Conformal Prediction in Predictive Modeling. A Transparent and Flexible Alternative to Applicability Domain Determination

10.1 billion for screening every year,

Integrated decision support for assessing chemical liabilities.

2.6 billion for screening every 2 years, and

Towards interoperable and reproducible QSAR analyses: Exchange of datasets.

3.5 billion for screening according to recommendations from the U.S. Preventive Services Task Force. Caution These results depend on assumptions, not all of which are equally well-supported. Implication Dollars saved under less expensive programs could be used to screen women who are not being screened. The Editors The frequency and appropriate age to start mammography screening for the detection of breast cancer have been debated in the United States for decades. Controversy intensified after the U.S. Preventive Services Task Force (USPSTF) recommended a change to biennial mammography on the basis that both annual and biennial screening reduce mortality rates, but biennial screening also decreases the negative effects (1). However, in the United States, there has been resistance to reducing frequency or modifying the age range for mammography. The USPSTF guidelines conflict with professional organizations, such as the American Cancer Society, which recommend annual screening from age 40 years and continued regardless of a womans age as long as she does not have serious, chronic health problems. Given the broad population that mammography serves, it is important to consider the economic effect of the conflicting guidelines. In 2009, the USPSTF recommended biennial screening for women between the ages of 50 and 74 years, with consideration of screening women aged 40 to 49 years on a riskbenefit decision (2). The USPSTF recommendations are based on a rigorous review of screening trials and work from the Cancer Intervention and Surveillance Modeling Network investigators that demonstrated that there is little net benefit in increasing the frequency of mammography (3). The Cancer Intervention and Surveillance Modeling Network modeling is corroborated by evidence from the Breast Cancer Surveillance Consortium, showing that false-positive recall and biopsy rates are significantly lower in the setting of biennial screening but without a significant increase in detected later-stage cancer (4, 5). The USPSTF recommendations on frequency are now in alignment with most European countries, where many of the defining mammography trials were conducted, with the exceptions of the United Kingdom and Finland, which screen every 3 years (611). Screening in the United States is delivered locally or regionally and covered by myriad payer and health plan organizations. Thus, the total resources required or the cost-tradeoffs of different recommendations are currently unknown. This study was designed to inform the debate by estimating the lower bound of the aggregate annual cost of mammography screening in the U.S. population when current (2010) screening practices are compared with guideline-recommended screening strategies. Our findings should be valuable to women, clinicians, and health policymakers alike who are aware of the many conflicting guidelines. Methods Study Design To estimate the cost of mammography in the United States, we created a simulation model using mammography screening in 2010 as our base case. We then simulated 3 strategies (annual, biennial, and USPSTF) from the payer perspective. Analyses were done using R (R Foundation for Statistical Computing, Vienna, Austria). Table 1 shows the 4 screening strategies, 1 of which is an estimate of actual practice (18, 19). The other 3 standardize on the population screened (85%) but differ on the age at which to start and stop and the frequency at which to screen. The biennial strategy represents the European approach, the annual strategy reflects the American Cancer Society (20) (among others) recommendations, and the USPSTF strategy (17) represents a risk-based strategy for screening those younger than 50 years and older than 75 years on the basis of their 2009 recommendations. Table 1. Model Inputs and Formulas The final output of the model was the aggregate cost of mammography screening per year. The summation included the costs of mammography, computer-aided detection (CAD), and recalls and biopsies. A description of the modeling methods is available in the Supplement. Supplement. Detailed Description of the Cost Models and the Sensitivity Analyses Inputs and Variables Model inputs were attained from several sources, including the Breast Cancer Surveillance Consortium (21), an observational data set designed to reflect mammography practice as it is done in the community and to reflect the distribution of women in the United States who have mammography, and they are listed in Table 1(13, 15, 16, 2224). All input variables except costs were age-specific. The number of mammography screenings was calculated by determining the population of women at risk by using census data. To focus on screening as opposed to diagnostic or surveillance mammography, we limited the population of women at risk to those between the ages of 40 and 85 years and excluded the number of women diagnosed with breast cancer in the past 5 years, who should be receiving surveillance mammography. We used data from the Behavioral Risk Factor Surveillance System 2010 Survey, a telephone health survey conducted by the Centers for Disease Control and Prevention, to determine the frequency and percentage of women receiving mammography. We corrected for survey bias by using the correction suggested by Rothman and colleagues (25). Although the survey does not distinguish between screening and diagnostic, we excluded women younger than 40 years and older than 85 years and those with recent history of breast cancer to best estimate screening as opposed to diagnostic mammography screenings (25). In our base-case model of actual practice, we included women who reported receiving mammography in the past 1 to 5 years and estimated the number who would have been screened in 1 given year; otherwise, the simulated strategies only simulated women receiving mammography every 1 or 2 years. The simulated strategies modeled a targeted participation of 85%, a screening participation achieved in the past for cervical cancer screening (26). For the USPSTF strategy, we modeled 20% of women aged 40 to 50 years as high-risk. We then simulated biennial screening for this cohort on the basis of evidence that women aged 40 to 49 years with a 2-fold increased risk have similar harmsbenefit ratios from biennial screening as women aged 50 to 74 years with average risk (27). The USPSTF strategy also modeled screening women between the ages of 70 and 85 years who are healthy, defined as having fewer than 3 self-reported chronic conditions as reported by Medicare (28). The percentage of recalled mammograms was obtained from the Breast Cancer Surveillance Consortium, using mammography screening performance data from 1908447 examinations for 749597 women screened from 2001 to 2007. Following the Breast ImagingReporting and Data System manual (29), a recall was defined as an initial Breast ImagingReporting and Data System assessment of 0 (needs additional imaging evaluation), 4 (suspicious abnormality), 5 (highly suggestive of malignancy), or 3 (probably benign finding) if it was accompanied by a recommendation for immediate work-up. Separate estimates were computed for recall rates at first and subsequent mammography screenings (that is, prevalent and incident screenings) as well as stratified by frequency of screening, digital versus film mammography, and a womans age. The estimated costs of the modeled strategies include the cost of screening mammography and the subsequent recall costs. Costs for mammography and CAD were determined using 2010 national Medicare reimbursements rates (16). Recall costs were calculated from the DMIST (Digital Mammographic Imaging Screening Trial) results of work-up costs, including additional imaging and biopsies from false-positive and true-positive examination results (24). We adjusted DMIST recall costs proportional to the use of digital versus film mammography in 2010. We adjusted all cost data to 2010 U.S. dollars on the basis of inflation as estimated by the medical portion of the Consumer Price Index (30). Sensitivity Analysis We used Monte Carlo simulations to estimate the uncertainty of our total cost estimates and quantify the sensitivity of the output (total cost) to the model inputs (Supplement) (31). In the sensitivity analysis, all terms in the formulas were assumed to be independent and follow -distributions as detailed in the Supplement. Role of the Funding Source This study was funded by the University of California and Safeway Foundation. The funding source had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; or preparation, review, or approval of the manuscript. Results Our model simulated screening practices in 2010 and estimated the aggregate U.S. population cost of mammography per year. Three mammography screening strategies advocated by various professional societies with targeted participation rates were also simulated and yielded costs that ranged from