Karolinska Institutet
Browse

Prostate cancer : investigating diagnosis, treatment and outcomes

Download (1.56 MB)
thesis
posted on 2024-09-03, 03:07 authored by Axel MöllerAxel Möller

Background: Prostate cancer is difficult to study. The prostate is located so that it is hard to palpate and because of the dense pelvic bones it is hard to visualize using radiology such as computed tomography. The prostate increases in size with age and it can be irregular in shape and consistency without the presence of cancer. Prostate cancer can be aggressive and deadly on one hand, while on the other hand, it can also be clinically insignificant, never causing any issues for the patient. Even if ignoring the most benign prostate cancers, the natural course of most prostate cancers is still slow. This means that researchers need long follow up times, up to 20 years, to observe cancer specific mortality. The efficiency of any diagnostic test or treatment often needs to be evaluated with surrogate endpoints because of the indolent nature of the disease. Prostate cancer should ideally be diagnosed before symptoms arise, since prostate cancer, in general, does not cause symptoms until it has become locally advanced or developed into metastatic disease. These are some of the many challenges for clinicians and researchers working with prostate cancer patients. Several blood-based tests have been suggested to enhance prostate cancer testing. In a large screening study, the Stockholm3 test has been shown to reduce the number of men undergoing prostate biopsies and to lower the detection rate of low-grade cancers, while still maintaining the detection rate of ISUP Gleason Group 2 to 5 cancers. The Stockholm3 test was trained and validated on prostate biopsies, and it is unknown how well the test performs, compared to PSA and PSA-density (PSA/prostate volume), when the whole prostate is available for pathological analysis. There are studies indicating that patients with prostate cancer could benefit from the use of existing drugs and vaccines not intended for prostate cancer so called drug repurposing. Recent findings suggest that the oral cholera vaccine could be effective in treating inflammatory diseases and certain prevalent cancers. A Swedish register study showed that administering the cholera vaccine following a prostate cancer diagnosis led to a nearly 50% reduction in prostate cancer-specific mortality rates. The ideal scenario after a prostatectomy for prostate cancer is to have undetectable levels of PSA, usually defined as less than 0.1 ng/mL. Nonetheless, some patients exhibit measurable PSA levels above 0.1 ng/mL at their first follow-up post-radical prostatectomy, a condition known as PSA persistence. This condition suggests possible metastatic disease, benign prostate remnants, or incomplete surgical removal. For those whose PSA persistence results from non-radical surgery, adjuvant radiotherapy targeting the prostate area might be beneficial.

Aims: Study 1. We aimed to validate the diagnostic performance of the Stockholm3 test in an independent, clinical practice cohort. Study 2. We aimed to investigate if oral cholera vaccine had an effect on overall mortality for men diagnosed with prostate cancer. Study 3. We aimed to evaluate and compare the effectiveness of PSA and PSAdensity against the Stockholm3 test in predicting the occurrence of clinically significant prostate cancer following prostatectomy. Study 4. We aimed to investigate the long-term prognosis of patients with PSA persistence, which risk factors that predict PSA persistence. As a secondary objective we aimed to assess the impact of adjuvant radiotherapy for men with PSA persistence.

Methods: Study 1. The study involved men, aged 45–75 who had never been diagnosed with prostate cancer and were scheduled for prostate biopsies at one of three clinics in Norway and Sweden. Before undergoing systematic prostate biopsies, these men had their blood drawn for Stockholm3 testing. Clinically significant prostate cancer was defined as ISUP Grade Group (GG) 2 or above. We compared the predictive performance for prostate cancer of the Stockholm3 test, PSA and PSA-density by modelling receiver operating characteristics curves and calculating the area under the curves. Study 2. We identified men diagnosed with prostate cancer using the Stockholm PSA and Biopsy Register, which is a population-based register of men from Stockholm, Sweden, with detailed social and prostate cancer data. Using the Swedish Prescribed Drug Register and combining it with the register of a large private vaccine clinic we could identify men who were exposed to vaccine after their prostate cancer diagnosis. To assess the effect of oral cholera vaccine on overall mortality we used a marginal structural Cox regression model. We made the same analysis for several other travel vaccines and malaria prophylaxis to use as a comparison. Study 3. The initial STHLM3 study included 58,818 men who were included in a screening by invitation of which 755 men underwent prostatectomy. These were included in the study. We defined clinically significant prostate cancer after prostatectomy as final pathology ISUP grade 2 or more, pT-stage 3 or more or tumor volume more than 0.5 cm3. We calculated the area under the curve (AUC) for predicting clinically significant prostate cancer at whole prostate specimen pathology for Stockholm3, PSA and PSA-density. Study 4. Using the Stockholm PSA and Biopsy Register, used in Study 2, we identified all men in Stockholm, Sweden who underwent a prostatectomy from 2004 to 2018. PSA persistence was defined as a PSA of >0.1 ng/mL at first blood test after 4 weeks or more postoperatively. Adjuvant radiotherapy was defined as radiotherapy within the first year after prostatectomy. Our endpoints were all cause mortality, prostate cancer specific mortality and disease progression, defined as hormonal treatment. Statistics used were chi-square test, Mann- Whitney U test, multivariate Cox regression and competing hazards regression.

Results: Study 1. Out of 533 participants, 263 (49%) were diagnosed with prostate cancer, with 162 of these men exhibiting ISUP Gleason Group (GG) ≥ 2 cancer. The Stockholm3 test showed better diagnostic effectiveness in identifying GG ≥ 2 prostate cancer compared to PSA and PSA-density. The predictive accuracy (area under the curve or AUC) for high-grade tumors was 64.2 for PSA alone, 74.8 for PSA density, and 85.9 for the Stockholm3 test. By applying a 10% risk threshold on the Stockholm3 test for when to biopsy, 38% of biopsies could be avoided, albeit with a delayed diagnosis in 6% (10 individuals) of cases with GG ≥ 2 cancer. Conversely, a biopsy threshold set at a PSA-density of 0.1 saved 35% of biopsies but resulted in a delayed diagnosis in 16% (26 individuals) of the men with GG ≥ 2 cancer. Study 2. We included 22,738 men with prostate cancer in the study, of which 4,248 (19%) died during follow-up. The median follow-up time was 5 years. 717 men received oral cholera vaccine after their prostate cancer diagnosis of which 36 died during follow-up. Receiving oral cholera vaccine after prostate cancer diagnosis was associated with a 54% reduced risk of dying (HR 0.46, p-value 0.0001). However, the same effect, or better, was seen for several of the other travel vaccines and for malaria prophylaxis. Study 3. We included 755 patients from the initial STHLM3 study that underwent prostatectomy. Postoperatively 93% had clinically significant prostate cancer on final pathology. For the whole study cohort, the AUC for clinically significant prostate cancer on whole prostate specimen was 77.3 (71.3-82.7) for Stockholm3, 70.5 (63.7-77.2) for PSA and 74.3 (68.3-80.3) for PSA-density. For men with diagnostic ISUP GG1 the AUC for clinically significant prostate cancer was 78.0 (68.1 -88.0) for Stockholm3, 65.3 (53.4-77.2) for PSA and 70.3 (59.3- 81.3) for PSA-density. Study 4. We included 8,905 men who underwent prostatectomy, among whom 496 (6%) exhibited PSA persistence post-surgery. Significant predictors of PSA persistence included the Gleason score, clinical T-status, and elevated preoperative PSA levels. Men experiencing PSA persistence showed increased rates of all-cause mortality (HR 2.23, 95% CI 1.60-3.12), cancer-specific mortality (HR 5.32, 95% CI 2.73-10.37), and an elevated risk of cancer progression (HR 6.44, 95% CI 5.40-7.68). Of these, 95 men with PSA persistence underwent adjuvant radiotherapy. While adjuvant radiotherapy was linked to a reduced risk of disease progression, the association was not statistically significant (SHR 0.53, 95% CI 0.14-1.96).

Conclusion: Study 1. The Stockholm3 test was independently validated in a clinical practice cohort and reduces the number of unnecessary biopsies while delaying diagnosis for a limited number of men. Study 2. We conclude that the effect of oral cholera vaccine seen on overall mortality is likely due to a healthy traveler bias and caused by residual confounding. Study 3. Stockholm3 and PSA-density are better predictors for clinically significant prostate cancer after prostatectomy compared to PSA. This validates the Stockholm3 test even further. Stockholm3 has the potential to be used for selecting men for active monitoring or treatment, but further studies are needed. Study 4. Men with PSA persistence following a prostatectomy demonstrated worse prognosis and more adverse tumor characteristics than those with undetectable PSA levels at first control after surgery. Additionally, adjuvant radiotherapy did not significantly impact disease progression, cancer-specific mortality, or overall mortality in these patients.

List of scientific papers

I. A. Möller, H. Olsson, H. Grönberg, M. Eklund, M. Aly, T. Nordström. The Stockholm3 blood-test predicts clinically significant cancer on biopsy: Independent validation in a multi-center community cohort. Prostate Cancer and Prostatic Diseases. 2019, 22, 137-142.
https://doi.org/10.1038/s41391-018-0082-5

II. A. Möller, K. Schwamborn, A. Spillmann, J. Hoogstraate, R. Szulkin, O. Akre, L. Egevad, M. Clements, M. Aly. Travel vaccines are strongly associated to reduced mortality in prostate cancer patients - a real effect or residual confounding? Vaccine. 2022, 40, 3797-3801.
https://doi.org/10.1016/j.vaccine.2022.05.028

III. A. Möller, T. Palsdottir, M. Eklund, T. Nordström, L. Egevad, Markus Aly. The performance of Stockholm3, PSA-density and PSA for predicting clinically significant prostate cancer after prostatectomy. [Submitted]

IV. A. Möller, O. Akre, T. Nordström, M. Eklund, L. Egevad, M. Aly. PSA persistence after prostatectomy – predictors, prognosis and the effect of adjuvant radiotherapy. [Manuscript]

History

Defence date

2024-02-23

Department

  • Department of Molecular Medicine and Surgery

Publisher/Institution

Karolinska Institutet

Main supervisor

Aly, Markus

Co-supervisors

Akre, Olof; Nordström, Tobias; Eklund, Martin

Publication year

2024

Thesis type

  • Doctoral thesis

ISBN

978-91-8017-252-3

Number of supporting papers

4

Language

  • eng

Original publication date

2024-01-19

Author name in thesis

Möller, Axel

Original department name

Department of Molecular Medicine and Surgery

Place of publication

Stockholm

Usage metrics

    Theses

    Categories

    No categories selected

    Keywords

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC