PSA-based machine learning model improves prostate cancer risk stratification in a screening population

Abstract

Context

The majority of prostate cancer diagnoses are facilitated by testing serum Prostate Specific Antigen (PSA) levels. Despite this, there are limitations to the diagnostic accuracy of PSA. Consideration of patient demographic factors and biochemical adjuncts to PSA may improve prostate cancer risk stratification. We aimed to develop a contemporary, accurate and cost-effective model based on objective measures to improve the accuracy of prostate cancer risk stratification.

Methods

Data were collated from a local institution and combined with patient data retrieved from the Prostate, Lung, Colorectal and Ovarian Cancer screening Trial (PLCO) database. Using a dataset of 4548 patients, a machine learning model was developed and trained using PSA, free-PSA, age and free-PSA to total PSA (FTR) ratio.

Results

The model was trained on a dataset involving 3638 patients and was then tested on a separate set of 910 patients. The model improved prediction for prostate cancer (AUC 0.72) compared to PSA alone (AUC 0.63), age (AUC 0.52), free-PSA (AUC 0.50) and FTR alone (AUC 0.65). When an operating point is chosen such that the sensitivity of the model is 80% the specificity of the model is 45.3%. The benefit in AUC secondary to the model was related to sample size, with AUC of 0.64 observed when a subset of the cohort was assessed.

Conclusions

Development of a dense neural network model improved the diagnostic accuracy in screening for prostate cancer. These results demonstrate an additional utility of machine learning methods in prostate cancer risk stratification when using biochemical parameters.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2

References

  1. 1.

    Siegel RL, Miller KD, Jemal A (2020) Cancer statistics, 2020. CA Cancer J Clin 70:7–30

    Article  Google Scholar 

  2. 2.

    Cabarkapa S, Perera M, McGrath S, Lawrentschuk N (2016) Prostate cancer screening with prostate-specific antigen: a guide to the guidelines. Prostate Int 4:125–129

    Article  Google Scholar 

  3. 3.

    Hugosson J, Roobol MJ, Mansson M et al (2019) A 16-yr follow-up of the european randomized study of screening for prostate cancer. Eur Urol 76:43–51

    Article  Google Scholar 

  4. 4.

    Pinsky PF, Prorok PC, Yu K et al (2017) Extended mortality results for prostate cancer screening in the PLCO trial with median follow-up of 15 years. Cancer 15(123):592–599

    Article  Google Scholar 

  5. 5.

    Toner L, Papa N, Perera M et al (2017) Multiparametric magnetic resonance imaging for prostate cancer-a comparative study including radical prostatectomy specimens. World J Urol 35:935–941

    CAS  Article  Google Scholar 

  6. 6.

    Ahmed HU, El-Shater Bosaily A, Brown LC et al (2017) Diagnostic accuracy of multi-parametric MRI and TRUS biopsy in prostate cancer (PROMIS): a paired validating confirmatory study. Lancet 25(389):815–822

    Article  Google Scholar 

  7. 7.

    Kasivisvanathan V, Emberton M, Moore CM (2018) MRI-targeted biopsy for prostate-cancer diagnosis. N Engl J Med 9(379):589–590

    Google Scholar 

  8. 8.

    Ito K, Yamamoto T, Ohi M, Kurokawa K, Suzuki K, Yamanaka H (2003) Free/total PSA ratio is a powerful predictor of future prostate cancer morbidity in men with initial PSA levels of 4.1 to 10.0 ng/mL. Urology 61:760–764

    Article  Google Scholar 

  9. 9.

    Loeb S, Catalona WJ (2014) The prostate Health Index: a new test for the detection of prostate cancer. Ther Adv Urol 6:74–77

    Article  Google Scholar 

  10. 10.

    Nitta S, Tsutsumi M, Sakka S et al (2019) Machine learning methods can more efficiently predict prostate cancer compared with prostate-specific antigen density and prostate-specific antigen velocity. Prostate Int 7:114–118

    Article  Google Scholar 

  11. 11.

    Snow PB, Smith DS, Catalona WJ (1994) Artificial neural networks in the diagnosis and prognosis of prostate cancer: a pilot study. J Urol 152:1923–1926

    CAS  Article  Google Scholar 

  12. 12.

    Stephan C, Xu C, Cammann H et al (2007) Assay-specific artificial neural networks for five different PSA assays and populations with PSA 2–10 ng/ml in 4,480 men. World J Urol 25:95–103

    Article  Google Scholar 

  13. 13.

    Stephan C, Xu C, Finne P et al (2007) Comparison of two different artificial neural networks for prostate biopsy indication in two different patient populations. Urology 70:596–601

    Article  Google Scholar 

  14. 14.

    Thompson IM, Ankerst DP, Chi C et al (2005) Operating characteristics of prostate-specific antigen in men with an initial PSA level of 3.0 ng/ml or lower. JAMA 294:66–70

    CAS  Article  Google Scholar 

  15. 15.

    Goldenberg SL, Nir G, Salcudean SE (2019) A new era: artificial intelligence and machine learning in prostate cancer. Nat Rev Urol 16:391–403

    Article  Google Scholar 

  16. 16.

    Carter HB (2013) American Urological Association (AUA) guideline on prostate cancer detection: process and rationale. BJU Int 112:543–547

    Article  Google Scholar 

  17. 17.

    Grossman DC, Curry SJ, Owens DK et al (2018) Screening for prostate cancer: US preventive services task force recommendation statement. JAMA 8(319):1901–1913

    Google Scholar 

  18. 18.

    Australia PCF. PSA testing and early management of test-detected prostate cancer. 2019. cited https://www.prostate.org.au/media/612113/PSA-Testing-Guidelines.pdf

  19. 19.

    McGrath S, Christidis D, Perera M et al (2016) Prostate cancer biomarkers: are we hitting the mark? Prostate Int 4:130–135

    Article  Google Scholar 

  20. 20.

    Loeb S, Sokoll LJ, Broyles DL et al (2013) Prospective multicenter evaluation of the Beckman Coulter Prostate Health Index using WHO calibration. J Urol 189:1702–1706

    Article  Google Scholar 

  21. 21.

    Nordstrom T, Vickers A, Assel M, Lilja H, Gronberg H, Eklund M (2015) Comparison between the four-kallikrein panel and prostate health index for predicting prostate cancer. Eur Urol 68:139–146

    Article  Google Scholar 

  22. 22.

    Chun FK, Graefen M, Briganti A et al (2007) Initial biopsy outcome prediction–head-to-head comparison of a logistic regression-based nomogram versus artificial neural network. Eur Urol. 51:1236–1240

    Article  Google Scholar 

  23. 23.

    Cuocolo R, Cipullo MB, Stanzione A et al (2019) Machine learning applications in prostate cancer magnetic resonance imaging. Eur Radiol Exp 7(3):35

    Article  Google Scholar 

  24. 24.

    Yoo S, Gujrathi I, Haider MA, Khalvati F (2019) Prostate cancer detection using deep convolutional neural networks. Sci Rep 20(9):19518

    Article  Google Scholar 

  25. 25.

    Corfield J, Perera M, Bolton D, Lawrentschuk N (2018) (68)Ga-prostate specific membrane antigen (PSMA) positron emission tomography (PET) for primary staging of high-risk prostate cancer: a systematic review. World J Urol 36:519–527

    Article  Google Scholar 

  26. 26.

    Perera M, Papa N, Roberts M et al (2020) Gallium-68 prostate-specific membrane antigen positron emission tomography in advanced prostate cancer-updated diagnostic utility, sensitivity, specificity, and distribution of prostate-specific membrane antigen-avid lesions: a systematic review and meta-analysis. Eur Urol 77:403–417

    Article  Google Scholar 

Download references

Funding

None.

Author information

Affiliations

Authors

Contributions

MP: manuscript writing/editing. RM: data collection, data analysis. NP: manuscript writing/editing. GB: data collection, data analysis. AE: data collection, data analysis. LS: project development, supervision. PS: project development, supervision. ES: data collection, data analysis.

Corresponding author

Correspondence to Marlon Perera.

Ethics declarations

Conflict of interest

RM, GB, AE are employees of Maxwell Plus. LS, PS and ES hold financial interests in Maxwell Plus.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Perera, M., Mirchandani, R., Papa, N. et al. PSA-based machine learning model improves prostate cancer risk stratification in a screening population. World J Urol (2020). https://doi.org/10.1007/s00345-020-03392-9

Download citation

Keywords

  • Prostate cancer
  • Prostate-specific membrane antigen
  • Prostate cancer screening
  • Machine learning
  • Artificial intelligence