Using real world data to bridge the evidence gap left by prostate cancer screening trials

Abstract

Background and purpose

Population-based prostate cancer screening is currently not recommended in the UK because harms may outweigh benefits. Recent changes to the diagnostic pathway have improved safety and accuracy, but uncertainty remains as to how much they have shifted the screening harm to benefit ratio. Our work uses modelling and real-world data (RWD) to bridge this evidence gap.

Materials and methods

We analysed outcomes of men entering the current prostate cancer diagnostic pathway using RWD from two NHS registries covering 16 hospitals. To assess improvements, we compared current UK clinical practice outcomes with those reported in the Cluster Randomised Trial of PSA Testing for Prostate Cancer (CAP)/ProtecT trial and past UK clinical practice, and to a model built to represent expected outcomes from the UK current diagnostic pathway.

Results

Out of 10 000 men who underwent a prostate specific antigen (PSA) test, we estimated that the proportion of men with no cancer after a biopsy following a PSA test decreased from 9.46% in the pre-magnetic resonance imaging CAP/ProtecT pathway to 2.33% and 1.52% in the Rapid Assessment for Prostate Imaging and Diagnosis (RAPID) and South West of England RWD pathways, respectively. Clinically insignificant prostate cancer diagnoses decreased to 0.73% in RAPID and 0.83% in South West, while the proportion of men experiencing sepsis reduced from a historic 0.10% to 0.02% in RAPID. We estimated an increase in clinically significant prostate cancer diagnoses in the RWD pathways.

Conclusions

This analysis of high-quality RWD estimates a 79% decrease in harm from the diagnostic process when comparing the previous pathway to actual outcomes from current UK clinical practice. This confirms and quantifies the harm reduction delivered by new diagnostic techniques.

Highlights

• •This study provides the first quantitative estimates of reduced short-term harms after a PSA test.
• •Fewer men face unnecessary biopsies and sepsis in RWD pathways.
• •Fewer men are diagnosed with clinically insignificant prostate cancer.
• •There is an increase in clinically significant prostate cancer diagnoses in RWD pathways.
• •The current UK prostate cancer diagnostic pathway is safer and more accurate.

Introduction

Prostate cancer is the most common cancer in men in the UK, with >52 000 new diagnoses and 12 000 deaths from prostate cancer recorded annually. Despite the high incidence and mortality rates, screening using prostate-specific antigen (PSA) is a subject of ongoing debate.¹ Most countries have chosen to recommend against population screening for prostate cancer, despite the ERSPC trial showing a 21% prostate cancer-specific mortality reduction after 13 years of follow-up² and a 30% reduction in metastatic disease after 12 years,³ because results from that and other screening trials suggest that the harms may outweigh the benefits.⁴ While there is evidence advances in diagnostic techniques, such as multiparametric magnetic resonance imaging (mpMRI) and transperineal (TP) biopsy may reduce harms and improve diagnostic accuracy,^5,6 it is currently unclear the extent to which those changes have shifted the harm-benefit ratio of screening, as all trials with long term follow up were conducted before the introduction of these improvements.

Three major trials have examined the effect of screening on prostate cancer-specific mortality: The European Randomised Study of Screening for Prostate Cancer (ERSPC), the Cluster Randomised Trial of PSA Testing for Prostate Cancer (CAP), and the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO). These trials applied the pre mpMRI prostate cancer diagnostic pathway, which involved a PSA test, followed by a standard Transrectal Ultrasound (TRUS) prostate biopsy if PSA levels were raised (generally ≥3 ng/ml, although some centres had a threshold of ≥2.5 ng/ml or ≥4 ng/ml).

Following this pathway, ERSPC demonstrated a 21% relative risk reduction in prostate cancer mortality after 13 years.⁷ ERSPC was the first trial to show a significant benefit from screening with 2-4 year PSA intervals. In contrast, CAP found a one-time screening led to overdiagnosis of clinically insignificant cancers with a slight decrease in prostate cancer mortality after 15 years of follow up.⁸ The PLCO trial found yearly PSA screening led to no reduction in prostate cancer-specific mortality.⁹ However, the PLCO trial had similar rates of PSA testing in both the intervention and control groups during the study period and poor adherence to biopsy referrals.¹⁰ These factors are now recognised as invalidating the results from that trial.¹¹

The ERSPC trial showed that prostate cancer screening can reduce prostate cancer specific mortality. However, it also highlighted several significant harms associated with screening, including short term harms such as unnecessary biopsies, adverse event resulting from biopsies, and long-term harms, such as overdiagnosis and overtreatment of clinically insignificant disease. At 16 years of follow-up, 570 men need to be invited for screening and 18 need to be diagnosed to prevent one prostate cancer death.² Despite a decrease in excess incidence with longer follow-up periods, it remained substantial, at 41% after 13 years.⁷ Considering this and other evidence, the UK National Screening Committee concluded harms outweigh benefits and does not currently recommend organised population screening for prostate cancer.¹²

There have been substantial changes in the prostate cancer diagnostic pathway since those three screening trials concluded. In the UK, since 2019, the use of a pre-biopsy mpMRI is recommended by National Institute for Health and Care Excellence (NICE).¹³ This change in guidance was informed by the PROMIS and PRECISION trials, which showed using mpMRI before biopsy could reduce unnecessary biopsies and diagnosis of clinically insignificant prostate cancer.^5,6 Although Prostate mpMRI was first rolled out in London, it is now widely available in the UK, with >90% of trusts already reporting access to pre biopsy prostate mpMRI as part of their prostate cancer diagnostic process in 2019.¹⁴

Current UK practice for diagnosis of prostate cancer therefore differs significantly from the practice tested in previous randomised trials. NICE guidelines still recommend the PSA test as the initial test for diagnosing prostate cancer. These guidelines do not recommend the use of additional blood and/or urine biomarkers, or risk calculators as part of the prostate cancer diagnostic pathway.¹³ In the UK, men over the age of 50 years are entitled to a free PSA test through the National Health Service (NHS) once they have been counselled on the possible pros and cons. However, as the UK does not have an organised screening programme this must be initiated in primary care by a clinician responding reactively to a request from a patient. Elevated PSA levels trigger a referral for a pre-biopsy mpMRI scan. NICE guidelines recommend mpMRI results should be reported using a five point Likert scale. Men with a Likert score of ≥3 should be referred for an mpMRI guided prostate biopsy for further investigation, to determine the presence or absence of prostate cancer.¹³ The use of pre-biopsy mpMRI allows targeting of lesions and avoiding unnecessary biopsies in patients with negative results. In addition, mpMRI guided targeted biopsies, and the increased use of TP biopsy has reduced biopsy harms, including sepsis.¹⁵

Materials and methods

We sought to describe the outcomes of men going through the current prostate cancer diagnostic pathway using recent Real-World Data (RWD) collected as part of routine audit within two NHS registries including 16 hospitals. Specifically, we aimed to assess the impact of the pathway changes on the harm and benefit ratio of a PSA test.

We compared the RWD pathways to a scenario model constructed by pooling data from multiple prospective studies representing the current UK prostate cancer diagnostic pathway, and to pre-MRI data from the CAP/ProtecT clinical trials. We aimed to explore differences in the number and proportion of men experiencing harm from prostate cancer screening across three metrics considered as possible harmful outcomes of a PSA test decision (Table 1). We also conducted exploratory analyses to determine if there were differences in outcomes between a screening cohort (CAP/ProtecT) and a cohort identified through RWD from UK clinical practice in the pre-MRI era.

Table 1.

Harm metrics

Harm metric	Definition	Justification for metric
Diagnosis of clinically insignificant prostate cancer (Gleason grade 3 + 3)	Number and proportion of men diagnosed with indolent prostate cancer after going through all steps in the diagnostic pathway.	Men diagnosed with clinically insignificant prostate cancer are exposed to harms including overtreatment, physical side effects from unnecessary treatment and psychological side effects from active surveillance of a harmless cancer.
A biopsy in a man without prostate cancer	Number and proportion of men who go through all the steps in the prostate cancer diagnostic pathway and have a biopsy that is negative for prostate cancer.	Men who receive a negative biopsy go through an invasive procedure and the psychological and physical side effects associated with it.
Sepsis as a result of biopsy	Number and proportion of men who experience sepsis after biopsy.	Sepsis as a result of biopsy can cause serious complications including: death, systemic inflammation, organ failure, and long-term physical and psychological side effects.

In our analysis, the two RWD pathways and the current UK diagnostic pathway scenario model represent current UK clinical practice, whereas the two pre-MRI pathways constructed using data from CAP/ProtecT and past UK clinical practice represent the historical diagnostic pathway, which is no longer recommended for use in the UK but reflects the pathway used in the major screening studies that are still used to consider the benefits and harms of screening today. Comparing old practice to current practice is the focus for this analysis.

RWD pathways

We mapped the flow of 10 000 men going through the current UK prostate cancer diagnostic pathway who had an initial PSA test using two different data registries: The Rapid Assessment for Prostate Imaging and Diagnosis (RAPID) cancer registry, and The South West England Prostate Dashboard Study Group (SW England).

Data for the following steps in the pathway were included: raised PSA blood tests (>3 ng/ml); mpMRI triage eligibility; mpMRI results; biopsy types; side effects from biopsy; and biopsy results.

Comparison to current UK diagnostic pathway model

We modelled a scenario where 10 000 men have an initial PSA test and assessed differences in outcomes between the RWD pathways and the model. The model outlines the quantities and percentages of men progressing through each stage in the prostate cancer diagnostic pathway. In this study, we present the results of a scenario where 10 000 men are subjected to a PSA test.

Comparison to the pre-MRI diagnostic pathways

We compared the RWD pathways and the current UK diagnostic pathway model with the historical pre-MRI diagnostic pathway. Our main comparator was built with data from the CAP/ProtecT trials. We also conducted additional exploratory analysis to assess any major differences in outcomes between a pre-MRI era screening cohort (CAP/ProtecT) and a pre-MRI routine clinical practice cohort to ensure that effects seen in our primary analysis were attributable to pathway changes rather than difference between a screening and presenting cohort.

Data sources

The data sources that informed the pathways included in this study are described in detail in Supplementary Material, available at https://doi.org/10.1016/j.esmorw.2024.100073. An overview of these sources is provided below.

RWD pathways

two multi-hospital prospective registries including 16 UK hospitals were included in the audit: The RAPID cancer registry, and SW England data dashboard.

The RAPID prostate cancer registry is a multi-centre registry of all patients referred with suspected prostate cancer through the RAPID pathway and includes data on clinical, MRI and histopathological outcomes.

The SW England data dashboard is a multi-centre registry of all patients referred with suspected prostate cancer under the 2-week wait pathway in the South West of England.

4419 patients from three high-volume hospitals in West London with suspected prostate cancer between 2017 and 2022 (RAPID), and 12 242 patients from 13 hospital across South West England with suspected prostate cancer between 2021 and 2022 were included in the analysis.

Current UK diagnostic pathway model

To model the flow of men through the current UK prostate cancer diagnostic pathway, we used data from cohort studies reporting on current UK practice patterns according to the national prostate cancer audit. Studies that have changed UK clinical practice were included. The data sources that informed the current UK diagnostic pathway model are described in Supplementary Material S1, available at https://doi.org/10.1016/j.esmorw.2024.100073.

Pre-MRI diagnostic pathway (CAP/ProtecT)

We used data from the CAP and ProtecT trials as a benchmark for the diagnostic pathway in place before mpMRI (Supplementary Material S2, available at https://doi.org/10.1016/j.esmorw.2024.100073). The CAP and ProtecT trial results were chosen as a comparator due to matched reporting of outcome measures with those used in the current diagnostic pathway model. The ERSPC trial was excluded as a comparator as the reporting of outcomes did not match the data points required to construct the model.

Pre-MRI diagnostic pathway (UK routine clinical practice)

We also used data from pre-MRI studies reporting on past UK clinical practice cohorts to represent the flow of men that went through the pre-MRI prostate cancer diagnostic pathway in a non-screening setting. The data sources that informed this model are described in Supplementary Material S3, available at https://doi.org/10.1016/j.esmorw.2024.100073.

Data analysis

Data from RAPID and SW England audits were analysed to map the quantities and proportions of men progressing through each stage of the prostate cancer diagnostic pathway. We compared the RWD pathways to the current UK diagnostic pathway model and the pre-MRI pathways to assess differences in the number and proportion of men experiencing harm and benefit across the three metrics described in Table 1. To enable direct comparisons, the results from RAPID and the SW England audits were normalised to scale them to the same number of men (1449) who are referred to a pre biopsy mpMRI based on estimates from the current UK diagnostic pathway model, assuming a consistent 16.1% conversion rate from PSA to referral for suspected prostate cancer across all pathways (Supplementary Material S1, available at https://doi.org/10.1016/j.esmorw.2024.100073). To calculate the overall reduction in harm, we compared the number and proportion of men that experienced the three short term forms of harm measured in the study between the pre-MRI diagnostic pathway (UK routine clinical practice) and the RWD pathways.

Results

Current UK diagnostic pathway model

The current UK model is described in Figure 1. We estimate in a scenario where 10 000 men have a PSA test, 1610 (16.1%) would have a PSA level ≥3 ng/ml and therefore would be referred to further investigation. This estimate was informed by a Clinical Research Practice Datalink study that calculated raised PSA level rates in the UK general population.¹⁶

Figure 1.

Current UK diagnostic pathway model. mpMRI, multiparametric magnetic resonance imaging; PSA, prostate-specific antigen; TP, transperineal.

If 10 000 undergo a PSA test, our model estimates ∼1449 (14.5%) would have raised PSA levels and meet the criteria for pre-biopsy mpMRI, and 161 (1.6%) would have raised PSA levels and be referred straight to a biopsy due to mpMRI contraindications. It is estimated 381 (3.8%) men would be ruled out of biopsy after normal mpMRI results. 1068 (10.7%) men would receive suspicious mpMRI results and subsequently be referred for prostate biopsies following mpMRI triage assessment. In total, 1229 (12.2%) men would have a biopsy.

A total of 823 (8.2%) men going through the pathway would be administered a TRUS biopsy, and 406 (4.0%) a TP biopsy. Seven (0.07%) men are estimated to experience sepsis (2 after TP and 5 after TRUS biopsy). A further 467 (4.7%) would be diagnosed with clinically significant (Gleason grade ≥3 + 4) and 111 (1.1%) with clinically insignificant (Gleason grade 3 + 3) prostate cancer. In all, 651 (6.5%) would go through all the diagnostic procedures in the pathway including biopsy prior to being excluded from a prostate cancer diagnosis.

Pre-MRI diagnostic pathways

The pre-MRI diagnostic pathway based on data from the CAP/ProtecT trials is described in Figure 2. In a scenario where 10 000 men underwent a PSA test before the introduction of pre-biopsy mpMRI, we estimated that 1560 (15.6%) would have a TRUS biopsy and 10 (0.1%) would develop sepsis after biopsy. Additionally, 231 (2.3%) men would be diagnosed with clinically significant and 383 (3.8%) men with clinically insignificant prostate cancer. The remaining 946 (9.5%) would undergo PSA testing and biopsy before being excluded from prostate cancer diagnosis.

Figure 2.

Pre-MRI diagnostic pathway (CAP/ProtecT data). CAP, Cluster Randomised Trial of PSA Testing for Prostate Cancer; MRI, magnetic resonance imaging; PSA, prostate-specific antigen.

The CAP/ProtecT trials reported data from a screening cohort, which is not fully representative of past UK clinical practice. To assess differences in outcomes between a pre-MRI era screening cohort and a pre-MRI era routine clinical practice cohort, we mapped the flow of men going through each step of the pre-MRI diagnostic pathway in past UK routine clinical practice. The pre-MRI diagnostic pathway based on data from past UK routine clinical pathway is described in Figure 3.

Figure 3.

Pre-MRI diagnostic pathway (UK routine clinical practice data). MRI, magnetic resonance imaging; PSA, prostate-specific antigen.

We estimate that 1610 (16.1%) men with a raised PSA level would have a biopsy and 21 (0.2%) would develop sepsis after biopsy. 338 (3.4%) men would be diagnosed with clinically significant prostate cancer, and 330 (3.3%) with clinically insignificant prostate cancer. The remaining 934 (9.3%) would be excluded from a prostate cancer diagnosis.

UK RWD pathways

When comparing the UK RWD pathways with the UK current model and the pre-MRI pathway (CAP/ProtecT), harm across all metrics was estimated to be reduced (Table 2). We estimate that after the introduction of pre biopsy mpMRI and TP biopsy, for every 10 000 men who had a PSA test, an additional 121 (52.3%) were diagnosed with clinically significant prostate cancer in RAPID and 166 (71.9%) in SW England when compared to the pre-MRI model, and 310 (80.9%) fewer men were diagnosed with clinically insignificant prostate cancer in RAPID and 300 (78.2%) in the SW England pathways when compared to the pre-MRI pathway (CAP/ProtecT). Notably, fewer men were estimated to go through all the diagnostic tests, including biopsy, before being ruled out of a prostate cancer diagnosis in both RWD pathways when compared to the pre-MRI pathway (CAP/ProtecT), 713 (75.4%) fewer in RAPID and 794 (83.9%) fewer in SW England.

Table 2.

Comparison of flow of men across pathways if 10 000 have a PSA test

Step in the pathway	Pre-MRI diagnostic pathway (CAP/ProtecT)	Pre-MRI diagnostic pathway (UK routine clinical practice)	Current UK diagnostic pathway model	Current diagnostic pathway
				RAPID (London) RWD	South West England RWD
Men with a raised PSA level	1610	1610	1610	1610	1610
Men eligible for mpMRI	N/A	N/A	1449	1449	1449
Men who have a biopsy	1560	1610	1229	658	632
Men who suffer sepsis after biopsy (harm)	10	21	7	2	No data
Men who have a biopsy showing no cancer (harm)	946	934	651	233	152
Men who have a biopsy showing insignificant prostate cancer (harm)	383	330	111	73	83
Men who have a biopsy showing clinically significant prostate cancer	231	338	467	352	397

CAP, Cluster Randomised Trial of PSA Testing for Prostate Cancer; mpMRI, multiparametric magnetic resonance imaging; PSA, prostate-specific antigen; RAPID, Rapid Assessment for Prostate Imaging and Diagnosis; RWD, real-world data.

When comparing the UK RWD pathways with the UK current pathway model, we estimate fewer men were diagnosed with clinically significant cancer, 352 and 397 respectively, as compared with 467 men in the current pathway model, and fewer men were diagnosed with clinically insignificant prostate cancer, with 73 in RAPID and 83 in SW England, as compared with 111 in the current pathway model. Fewer men were estimated to require an arguably unnecessary biopsy, 233 in RAPID and 153 in SW England when compared with 651 in the current diagnostic pathway.

The comparison of the current UK diagnostic pathway model with RWD from UK routine clinical practice demonstrated greater potential reduction in harm across all three metrics of harm in RAPID and SW England audit data (Table 3). Out of the 10 000 men who had an initial PSA test, in RAPID, 2.3% of men had a biopsy showing no prostate cancer, 0.73% were diagnosed with clinically insignificant prostate cancer and just 0.02% experienced sepsis after biopsy. In SW England, 1.52% of men who had an initial PSA test had a biopsy but did not have prostate cancer, and 0.83% were diagnosed with clinically insignificant cancer.

Table 3.

Proportion of men experiencing harm outcomes after a PSA test across pathways

Possible harm outcome after a PSA test	Pre-MRI diagnostic pathway (CAP/ProtecT)	Pre-MRI diagnostic pathway (UK routine clinical practice)	Current UK diagnostic pathway model	Current diagnostic pathway
				RAPID (London)RWD	South West England RWD
% of men having a PSA test who have a biopsy showing no cancer	9.46	9.34	6.51	2.33	1.52
% of men having a PSA test who have a biopsy showing insignificant cancer	3.83	3.30	1.11	0.73	0.83
% of men having a PSA test who suffer sepsis	0.10	0.21	0.07	0.02	No data

CAP, Cluster Randomised Trial of PSA Testing for Prostate Cancer; MRI, c magnetic resonance imaging; PSA, prostate-specific antigen; RAPID, Rapid Assessment for Prostate Imaging and Diagnosis.

When comparing outcomes of men going through the CAP/ProtecT pre-MRI pathway and the past UK routine clinical practice pre-MRI pathway (Table 3), we found that more men experienced sepsis after biopsy (0.1% in the CAP/ProtecT pathway and 0.2% in the past UK clinical practice pathway). There were slightly fewer clinically insignificant diagnoses, and slightly more clinically significant diagnoses in the pre-MRI UK routine clinical practice diagnostic pathway when compared to the pre-MRI CAP/ProtecT pathway, but no substantial differences were noted giving confidence that it is valid to compare the CAP/ProtecT cohort to our modern practice Real World cohorts to establish harm reduction through pathway changes.

Discussion

Key findings in this study suggest that since the addition of pre-biopsy mpMRI, mpMRI guided biopsy, and TP biopsy into the prostate cancer diagnostic pathway, the short-term harms associated with prostate cancer screening have been notably reduced across all three major forms of harm potentially resulting from a PSA test. We estimate a decrease in the number of men exposed to unnecessary biopsy, a reduction in cases of sepsis following biopsy, and a decrease in overdiagnosis of clinically insignificant prostate cancer. Additionally, we report an increase in the detection of clinically significant prostate cancer when comparing outcomes between 10 000 men undergoing PSA testing using the pre-MRI pathway (CAP/ProtecT) versus the current UK diagnostic pathway model and the UK RWD pathways. Importantly, the harm reduction in real practice is even greater than that predicted by our model. We hypothesise routine clinical practice has continued to improve since the reports of pivotal clinical trials, leading to improved outcomes in RWD. However, differences observed between the UK RWD pathways and the current UK diagnostic pathway model could be due to differences in the characteristics of men enrolled in clinical trials and men in UK clinical practice, as well as differences in access to PSA testing and PSA testing patterns between groups.

The PROMIS and PRECISION studies demonstrated mpMRI could reduce unnecessary biopsy by 27% while also increasing the detection of clinically significant cancers and reducing diagnosis of 3 + 3 prostate cancers compared to TRUS biopsy alone.^5,6 Data from RAPID and SW England audits suggest the outcomes of PROMIS and PRECISION are being replicated and improved in UK routine clinical practice. It is also noteworthy that in the RAPID audit, only 0.02% of patients who undergo the diagnostic pathway develop sepsis. This percentage is considerably lower compared to the reported incidence of 0.1% in CAP,¹⁷ and estimates from the current pathway model (0.07%). This is likely attributable to the increased use of TP biopsy in the UK, which is associated with fewer side effects.¹⁵ While TRUS biopsy remains common, adoption of TP biopsy is increasing, with the National Prostate Cancer Audit recently reporting that in 2021, 40% and 24% of men diagnosed with prostate cancer undergo a TP biopsy in England and Wales respectively, with the remainder diagnosed via TRUS biopsy.¹⁸ The RAPID and SW England audits include data from 16 hospitals, all of which follow the prostate cancer diagnostic pathway recommended by NICE guidelines. The availability of MRI services in these hospitals is reflective of MRI availability across the UK.

This analysis of high-quality real-world evidence estimates a 79% decrease in harm from the diagnostic process when comparing outcomes in the pre-MRI routine clinical practice pathway to those in the current UK RWD pathways (Rapid and SW England pathways).

Strengths

Our study provides, to our knowledge, the first quantitative estimates of reduction in harms and comparative analysis with real-world outcomes from UK routine clinical practice. This study includes high quality RWD from 16 UK hospitals, including >16 000 consecutive patients. While acknowledging the limitations inherent in RWD in comparison to evidence from large scale randomised clinical trials we believe that, in this case, RWD provides an accurate picture of actual outcomes for men being tested for prostate cancer in the UK today. As such it is arguably a more accurate picture of the real likelihood of suffering a harmful outcome than the evidence provided by any of the three major screening trials previously conducted due to the pathway changes that have occurred since.

Limitations

This scenario analysis estimated the reduction in the immediate harms that may result from having a PSA blood test. However, we were unable to assess, within the same analysis, impacts on downstream overtreatment, treatment choices, and quality of life. Evidence from the National Prostate Cancer Audit suggests that >95% of men diagnosed with very low risk (Gleason grade 3 + 3, T1) localised prostate cancer opt for active surveillance as initial treatment in England and Wales,¹⁹ but the proportion that adhere to active surveillance is unknown.²⁰ The high uptake of active surveillance has potentially reduced overtreatment of very low risk prostate cancer, but retrospective data makes measuring changes in overtreatment among men diagnosed difficult.²¹ Additionally, we were unable to quantify changes in the long-term benefits of screening, such as reduction in metastatic progression and prostate cancer-specific mortality. These end points need to be further tested in randomised control trials.

One of the harms associated with diagnosis of a clinically insignificant cancer relies on the current consensus (largely driven by results from the ProtecT trial) that prostate cancers with no Gleason pattern 4 evident on biopsy is indolent.^22,23 This is reflected in NICE clinical guidelines which recommend active surveillance for men with those low risk (Gleason 3 + 3 or Gleason Grade Group 1) cancers.¹³ However, it has been suggested that increased accuracy enabled by MRI-targeted biopsy may have resulted in a grade shift with cancers that were classified as 3 + 3 pre-MRI era now more likely to be correctly called as having some pattern 4 cancer on biopsy, but the oncological risk of these cancers is unclear.²⁴ It is possible that some of the reduction in diagnosis of clinically insignificant cancer noted in our analysis of current RWD may be driven, not by avoidance of detection of indolent cancers, but a reclassification effect as they are now more accurately diagnosed. Importantly, however, the other two harm measures (sepsis and negative biopsies) are objective, hard endpoints and so do not suffer from this potential confounding factor. The reduction in unnecessary biopsies alone will have removed significant harm from the overall diagnostic pathway.

Our primary analysis compares RWD from current UK routine clinical practice, representing the current prostate cancer diagnostic pathway to pre-MRI data from CAP/ProtecT, representing the older prostate cancer pathway used during the evaluation of screening effects in the UK. We recognise that this comparison may potentially introduce bias, as we are comparing a screening cohort with a cohort of men presenting in UK clinical practice without a screening invitation. To address this potential limitation, we conducted an additional analysis using multiple data sources to measure outcomes from a cohort of men presenting without screening in UK routine clinical practice in the pre-MRI era. Effectively that additional cohort allows us to compare outcomes from the screened pre-MRI group (from CAP/ProtecT) to a pre-MRI presenting without screening group equivalent to our current practice real world cohorts but in the previous time period. We found no major differences between the screened and presenting (trial or practice) groups in the pre-MRI era so are confident that the primary comparison is valid and differences seen are due to changes in the pathway rather than changes in the characteristics of the men presenting through different routes. Our study compares the diagnostic techniques in place in the past in the UK with current practice.

Our research focuses on two audits that incorporate 16 UK hospitals. Although pre-biopsy mpMRI is widely available across the UK and recommended by guidelines,¹³ there is variation between trusts, it is not certain that the use of mpMRI results to rule men out of biopsy is always happening to best practice standards at an individual (as opposed to NHS Trust) level.

We did not have access to the baseline characteristics of the patients included in our study. Lack of data on ethnicity and family history meant we were unable to model the current UK diagnostic pathway for men at higher risk of prostate cancer. Black men enter the pathway with double the risk of prostate cancer diagnosis and death when compared with other men, and men with a first degree relative with prostate cancer are 2.5 times more likely to be diagnosed when compared to men without a family history.²⁵ Thus, they could potentially derive double the benefit without additional risk. However, the lack of data means we are unable to add weight to that assumption through this work.

Conclusions

By estimating the extent of harm reduction and assessing the possible impact of the translation of predicted outcomes to clinical practice, this work represents an important step toward re-evaluating the harm-to-benefit ratio of prostate cancer screening in the modern era. Our findings are based on high-quality RWD and suggest that since the introduction of new technologies into the prostate cancer pathway, it has become safer and more effective. These findings provide pertinent evidence regarding possible pathway changes for prostate cancer testing in the UK in the near term and adds to the evidence base to support decision making about how to diagnose prostate cancer, including decisions about introducing screening and more proactively offering PSA blood tests to men at higher risk. There is wide consensus that men should enter this pathway only in possession of true, accurate and up-to-date information about benefits and harms that they may encounter. The results from this study can be used as the basis for the information provided to men at risk of prostate cancer considering entering the diagnostic pathway for the first time, addressing the evidence gap left by pre-MRI prostate cancer screening trials.

Nevertheless, we acknowledge this study does not provide level 1 evidence, which can only come from large scale randomised controlled trials designed and powered to assess the long-term outcomes alongside the immediate harms from diagnosis such as the recently announced TRANSFORM trial in the UK.²⁶ It is essential that those long-term studies are funded and delivered to test the current, as well as new, pathways that might further reduce harms, improve accuracy and specificity for aggressive and lethal prostate cancer, and reduce false negative results at all stages of the pathway. However, given that completion of these studies will likely take several years, this research provides valuable insights that should be taken into consideration while waiting for longer term results.

Further work should also explore whether there are cohorts of men at highest risk, such as black men who face double the risk of both diagnosis and death compared to other men,²⁷ for whom the balance on harms and benefits could have more clearly shifted in favour of proactive testing.