Genetic risk score could improve triage of men with suspected prostate cancer

19th August 2022

A higher genetic risk score in symptomatic men is significantly associated with the development of prostate cancer over the next 2 years

Use of a genetic risk score (GRS) in men with symptoms suggestive of prostate cancer could enable the identification of those at risk and fast track them for further investigation according to the results of a study by UK researchers using a cohort from the UK Biobank database.

Prostate cancer was responsible for 1.4 million global cases in 2020 and is the 2nd most commonly occurring cancer in men. Screening for the cancer may help to save lives and a European study which followed-up on men for 13 years found that one prostate cancer death was averted per 781 men invited for screening although the authors concluded that ‘despite our findings, further quantification of harms and their reduction are still considered a prerequisite for the introduction of populated-based screening.’

Symptoms potentially suggestive of prostate cancer include lower urinary tract symptoms (LUTS) including nocturia, urinary frequency or poor stream but these are often present at the time of a prostate cancer diagnosis.

The use of prostate-specific antigen (PSA) in men with LUTS is another means of screening though a 2022 systematic review concluded that the available evidence suggests PSA is highly sensitive but poorly specific for prostate cancer detection in symptomatic patients. 

Prostate cancer is a highly heritable disease and there are 269 known genetic risk variants such that the use of a genetic risk score offers an approach for personalised risk prediction.

Moreover, it has been suggested that the use of a GRS provides additional information to improve upon current practices in prostate cancer screening by risk-stratifying patients before initial prostate-specific antigen testing.

However, to date, GRS is not routinely used and for the present study, researchers set out to assess whether a GRS was able to predict a new diagnosis of prostate cancer in men with LUTS over the next 2 years.

The team used information held in the UK Biobank and included men with LUTS but no recorded diagnosis of prostate cancer and a GRS was calculated for each participant based on the known genetic variants.

The association between the GRS and prostate cancer diagnosis within 2 years of symptom onset was calculated using logistic regression and GRS scores were split into quintiles for analysis.

Genetic risk score and prostate cancer

A total of 6777 men with prostate cancer symptoms were included and of whom, 247 men had a record of prostate cancer within 2 years. For the remaining 6530 men, 62 died during the 2-year follow-up period, leaving 6448 as control patients.

Among men with symptoms, the GRS was associated with the development of symptoms over the next 2-years (odds ratio, OR = 2.12, 95% CI 1.86 – 2.41). The risk of prostate cancer was also age dependent such that men in the highest GRS quintile had a 13% greater incidence than those in the lowest (2.3%).

When researchers added age to the GRS score, the predictive power of the model improved with an area under (AUC) the receiver operating characteristic curve of 0.77 (95% CI 0.74 – 0.80) which was higher than either the GRS (AUC = 0.70) or age (AUC = 0.68) alone.

The authors discussed how the results of the study could be used to enable identification of men at low risk of prostate cancer and therefore avoid further testing but also to fast track those with the highest risk.

A recognised limitation of the study was that genetic sequencing is not currently available in the UK so that the GRS approach cannot be implemented at present.

Citation
Green HD et al. Applying a genetic risk score for prostate cancer to men with lower urinary tract symptoms in primary care to predict prostate cancer diagnosis: a cohort study in the UK Biobank Br J Cancer 2022

Higher body fat levels in men linked to increased risk of prostate cancer death

12th May 2022

Higher body fat levels in men leads to an increased risk of prostate cancer death according to a meta-analysis of prospective studies

A higher body fat level in men is associated with an elevated risk of prostate cancer death according to a meta-analysis of prospective studies by researchers from the Nuffield Department of Population Health, Cancer Epidemiology Unit, University of Oxford, Oxford, UK.

Prostate cancer is the second most frequent cancer diagnosis made in men and the fifth leading cause of death worldwide and in 2020 there were more than 1.4 million new cases of prostate cancer.

Prior evidence indicates that there is a positive association between height and the risk of prostate cancer, with taller men being at a greater risk but also that those with greater adiposity, have an elevated risk of high-grade prostate cancer and prostate cancer death.

Moreover, other work suggests that a higher body fat level, based on central adiposity is a more relevant factor and that a higher waist circumference was an important risk factor for prostate cancer.

For the present study, the Oxford team use data from the UK Biobank and focused on men who had originally undergone anthropometric measurements (e.g., height, weight, waist and hip circumference).

A subgroup of these men also underwent abdominal MRI and a dual-energy X-ray absorptiometry (DXA) scan and for whom body mass index (BMI), waist and hip circumferences were re-assessed.

The primary outcome of interest was prostate cancer as the underlying cause of death. In addition, the researchers combined their Biobank data with other published prospective studies to undertake a dose response meta-analysis.

Higher body fat levels and prostate cancer death

Among a cohort of 21,8237 men with a mean age at recruitment of 56.5 years, over a follow-up period of 11.6 years, 661 men (mean age = 63.1 years), died of prostate cancer.

In a multivariable-adjusted model, there was no statistically significant association of BMI, body fat percentage and waist circumference and prostate cancer mortality. However, for the waist to hip ratio (WHR), this association was significant per 0.05 unit increase (hazard ratio, HR = 1.07, 95% CI 1.01 – 1.14, P for trend = 0.028) when comparing the highest to lowest WHR quartiles.

In the meta-analysis, the hazard ratio was 1.10 (95% CI 1.07 – 1.12) for every 5kg/m2 increase in BMI, 1.03 for every 5% increase in body fat percentage, and 1.06 for every 0.05 increase in WHR.

Using the estimate for the effect of BMI from the meta-analysis, the authors estimated that as approximately 11,900 men died from prostate cancer each year (averaged between 2016 – 2018) and if their estimate was accurate, a reduction in mean BMI of 5kg/m2 would potentially lead to 1309 fewer prostate cancer deaths every year in the UK.

They concluded that men with higher body fat (both total and central) were at a higher risk of death from prostate cancer and that these findings provided a reason for men to maintain a healthy weight.

Citation
Perez‐Cornago A et al. Adiposity and risk of prostate cancer death: a prospective analysis in UK Biobank and meta-analysis of published studies BMC Med 2022

Multi-parametric ultrasound diagnostic performance similar to MRI for prostate cancer

11th March 2022

Multi-parametric ultrasound is broadly similar to MRI for detecting prostate cancer but combing both would increase rates cancer detection

Multi-parametric ultrasound is a useful alternative to magnetic resonance imaging (MRI) for the detection of prostate cancer although detection rates would be improved by using both imaging modalities. This was the conclusion of a comparative study by a team from the Division of Surgical and Interventional Sciences, University College London, London, UK.

Prostate cancer has been found to be the second most frequent cancer diagnosis made in men and the fifth leading cause of death worldwide.

According to the World Cancer Research Fund, there were 1.3 million new cases in 2018 and which resulted in 358,989 deaths (3.8% of all deaths caused by cancer in men) in 2018. Among men who present with an elevated prostate specific antigen (PSA) level or a palpable abnormality after a digital-rectal examination, there is an increased potential for the diagnosis of prostate cancer.

According to a 2019 Cochrane systematic review, the use of a magnetic resonance imaging pathway has the most favourable diagnostic accuracy in clinically significant prostate cancer detection. 

Nevertheless, multi-parametric MRI is not universally used and multi-parametric ultrasound appears to be a promising alternative approach with the available evidence confirming that combining different ultrasound scans significantly improves diagnostic performance in prostate cancer.

For the present study and given the widespread availability of multi-parametric ultrasound, the UK team sought to compare the overall level of agreement between ultrasound and MRI scanning for the diagnosis of clinically significant prostate cancer.

They undertook a prospective, multi-centre, paired-cohort study, in which patients with an elevated PSA or a digital-rectal examination abnormality, underwent both multi-parametric ultrasound (MPUS) and multi-parametric MRI (MPMRI). In cases where either modality identified positive findings, patients were referred for targeted biopsies.

The study had two primary endpoints: the proportion of positive results from both imaging tests and the level of agreement between them. The second outcome was the detection of clinically significant prostate cancer after biopsy defined as Gleason >4+3 of any length and/or maximum cancer core length of >6mm of any grade.

Multi-parametric ultrasound and cancer detection

A total of 370 men with a mean age of 64.5 years were included in the trial, of whom, 306 had both imaging scans and 257 underwent a subsequent prostate biospy.

MPUS was positive for 89 % of patients and MRMRI in 78% (difference 11.1%, 95% CI 5.1 – 17.1). In addition, the agreement in lesion detection between MPUS and MRMRI was 73.2%.

Any cancer was detected in 52% of the 257 patients with 32% being clinically significant. However, there were differences between the two modalities. Overall, MPUS detected 4.3% fewer clinically significant cancers than MRMRI although MPUS detected 7.2% (6/83) significant cancers missed by MRMRI but the latter detected 20.5% of significant cancers missed by MPUS.

Nevertheless, there was a 91.1% agreement between the two modalities on the detection of clinically significant cancer.

The authors determined that combining MPUS and MPMRI would have led to the detection of 99 clinically significant cancers. They added that while MPUS detected 4.3% few significant cancers and would have resulted in 11.3% more patients being referred for a biopsy, both imaging modalities had missed cancers detected by the other technique.

As a final point, they added that using both modalities would increase the detection of clinically significant cancers to either test alone.

Citation
Grey ADR et al. Multiparametric ultrasound versus multiparametric MRI to diagnose prostate cancer (CADMUS): a prospective, multicentre, paired-cohort, confirmatory study Lancet Oncol 2022

Darolutamide combination treatment increases overall survival in metastatic prostate cancer

28th February 2022

Darolutamide combined with androgen-deprivation therapy and docetaxel increased overall survival in men with metastatic prostate cancer.

Darolutamide combined with androgen-deprivation therapy and docetaxel, provides a better overall survival in men with metastatic, hormone-sensitive prostate cancer compared not using the drug.

This was the finding of a randomised trial by researchers from the Genitourinary Malignancies Program, Massachusetts General Hospital Cancer Center and Harvard Medical School, Boston, US.

Prostate cancer is the second most frequent malignancy in men worldwide and in 2018, there were 1,276,106 new cases and 358,989 deaths, representing 3.8% of all male cancer deaths.

Among men with metastatic, hormone-naive, prostate cancer the addition of an androgen-receptor pathway inhibitor such as abiraterone, apalutamide, enzalutamide or docetaxel improves overall survival compared to androgen-deprivation therapy alone.

Darolutamide is an androgen-receptor inhibitor and in the Phase III ARAMIS study, it was found that for men with non-metastatic, castration-resistant prostate cancer, darolutamide treatment, metastasis-free survival was significantly longer with darolutamide than with placebo.

However, whether combining darolutamide, androgen-deprivation therapy and docetaxel, would increase survival among patients with metastatic, hormone-sensitive prostate cancer is currently unclear and was the rational for the current study by the US team.

The researchers included adults (> 18 years of age) with an Eastern Cooperative Oncology Group performance status (ECOG) of 0 or 1 (where higher scores, reflect a greater disability). All received androgen-deprivation therapy, in the form of a luteinising hormone-releasing hormone or analogue) or underwent orchiectomy within 12 weeks before randomisation.

Participants were randomised 1:1 to darolutamide 600 mg twice daily or matching placebo and received six cycles of docetaxel with prednisolone within 6 weeks after randomisation.

The primary endpoint of the trial was overall survival, defined as the time from randomisation until any cause of death. Secondary outcomes included time to castration-resistant prostate cancer and time to pain progression and participants were assessed every 12 weeks.

Darolutamide and overall survival

A total of 1,305 men with metastatic disease and a median age of 67 years were included in the final dataset, with most (71.1%) having an ECOG score of 0.

The risk of death was 32.5% lower in the darolutamide group (hazard ratio, HR = 0.68, 95% CI 0.57 – 0.80, p < 0.001). After four years, the overall survival was 62.7% in the darolutamide group compared to 50.4% in the placebo arm.

With respect to the secondary outcomes, the time to castration-resistant prostate cancer was significantly lower in the darolutamide group (HR = 0.36, 95% CI 0.30 – 0.42, p < 0.001) as was the time to pain progression (HR = 0.79, 95% CI 0.66 – 0.95, p = 0.01).

In terms of safety, the incidence of adverse effects was similar in both groups (99.5 vs 98.9, darolutamide vs placebo) as was the incidence of a serious adverse event (44.8% vs 42.3%).

Based on these findings, the authors concluded that darolutamide combined with androgen-deprivation therapy and docetaxel in men with metastatic, hormone-sensitive prostate cancer, led to a significantly longer overall survival.

Citation
Smith MR et al. Darolutamide and Survival in Metastatic, Hormone-Sensitive Prostate Cancer New Engl J Med 2022.

Elevated lipoprotein A levels positively associated with higher risk of prostate cancer

4th February 2022

Elevated lipoprotein A levels have been found to be associated with an increased risk of total, advanced and early age onset prostate cancer

A study by researchers from the Department of Epidemiology and Biostatistics, Imperial College, London has revealed a positive association between lipoprotein A concentrations and the risk of total, advanced and early age onset prostate cancer.

Prostate cancer is the second most common cancer diagnosis in men and the fifth leading cause of death worldwide. The GLOBOCAN 2018 data revealed how prostate cancer accounted for 7.1% of all incident cancers and 3.8% of all deaths. Moreover, there is some degree of biological heterogeneity with prostate cancer such that while there are several traditional risk factors including, age, ethnicity, family history, smoking, alcohol consumption, vasectomy and diet, the risk are different for more advanced disease. For instance, with more aggressive prostate cancer, the risk factors have mirrored those associated with cardiovascular disease such as obesity, dyslipidaemia, glucose intolerance, metabolic syndrome, unhealthy dietary habits or caloric excess, lack of physical activity and inflammation. Nevertheless, there remains much uncertainty over the precise relationship between cardiovascular risk factors and prostate cancer. For instance, though dyslipidaemia, as characterised for example, by elevated cholesterol levels, is a potential risk factor, a meta-analysis of 14 studies concluded that blood total cholesterol, HDL and LDL levels were not associated with the risk of either overall prostate cancer or high-grade prostate cancer. Furthermore, treatment for hyperlipidaemia, especially with the use of statins found that although there was no evidence that cholesterol lowering is beneficial for the prevention of low-grade or localised prostate cancer, there did appear to be an association between statin use and a reduced risk of advanced or high-grade prostate cancer. Much of the evidence for the relationship between prostate cancer and lipids comes from observational studies and one approach to control for the various confounders present in such analyses, is the use of Mendelian randomisation. In fact, one such Mendelian randomisation study assessing whether circulating lipids causally influence prostate cancer risk, concluded that there was some weak evidence to suggest that higher LDL and triglyceride levels increase aggressive prostate cancer risk.

For the present study, the UK researchers sought to determine whether genetically predicted lipid traits were associated with the overall risk of prostate cancer. They used data from the UK Biobank and looked for associations between not only cholesterol, HDL and LDL levels and prostate cancer but also with other lipid sub-fractions including lipoprotein A, apolipoprotein A and B.

Lipoprotein A and prostate cancer

Univariate analysis revealed that HLD and LDL cholesterol levels, triglycerides and apolipoprotein A and B concentrations were not associated with total prostate cancer risk. There was also no significant association between these sub-fractions and advanced prostate cancer. However, the analysis did reveal that elevated lipoprotein A levels were associated with advanced prostate cancer (odds ratio, OR = 1.03, 95% CI 1.0 – 1.06, p = 0.046) and with an increased risk of early onset prostate cancer (OR = 1.25, 95% CI 1.107 – 1.42, p < 0.001). Using multivariate analysis, the authors found that lipoprotein A levels were associated with total (OR = 1.068, p = 0.034), advanced (OR = 1.07, p = 0.055) and early onset prostate cancer (OR = 1.15, p = 0.028).

The authors concluded that their data indicated a positive association between lipoprotein A levels and the risk of total, advanced and early onset prostate cancer. In addition, they suggested that screening for high lipoprotein A levels, could be used to identify high-risk groups for prostate cancer.

Citation
Ioannidou A et al. The relationship between lipoprotein A and other lipids with prostate cancer risk: A multivariable Mendelian randomisation study PLoS Med 2022

Dietary supplements used by 40% of those diagnosed with cancer

4th January 2022

Dietary supplements have been found to be used by 40% of adults diagnosed with breast, prostate and colorectal cancer

Dietary supplements (DS) are used by 40% of adult patients diagnosed with either breast, prostate or colorectal cancer according to research by a team from the Department of Behavioural Science and Health, University College London, UK.

Survival from cancer appears to be increasing, with a 2018 global surveillance study finding that survival trends are generally increasing, even for some of the more lethal cancers.

While evidence supporting various strategies aimed at reducing cancer risk in those living with and beyond cancer is rather limited, a 2018 report by the World Cancer Research fund and the American Institute for Cancer research, is clear in its view that ‘high-dose dietary supplements are not recommended for cancer prevention’, encouraging individuals to meet their nutritional needs through diet alone.

Nevertheless, some data shows that cancer survivors tend to report a higher usage of DS than those with the disease.

For the current study, the authors sought to gain a better understanding the range of and reasons for, use of DS among survivors of breast, prostate and colorectal cancer. They undertook a cross-sectional survey using data from the Advancing Survival Cancer Outcomes Trial (ASCOT) and asked respondents with each of the three cancers their thoughts about lifestyle and cancer, use of specific foods, e.g. fruits, vegetables, meat and high calorie foods together with information on the use of DS and any other non-prescribed treatments such as herbal extracts.

Respondents were asked to express their views (using a Likert scale) on the perceived importance of supplements as an approach to prevent cancer reoccurrence.

Findings

A total of 1049 participants with mean age of 64.4 years (62.1% female) provided usable data for analysis. Breast cancer was the most common (54.4%) among respondents, followed by prostate (25.2%) and colorectal (20.4%). In addition, the majority were of white ethnicity (94%) and 68% had either no (34.9%) or at least one co-morbidity.

In total, 40% of respondents reported DS use, of whom, 32% believed that these supplements were important for a reduction in cancer recurrence. The most commonly used form of supplements were fish oils (13.1%), followed by calcium and vitamin D (9.1%) and multivitamin and minerals (8.2%).

Using regression analysis, the only factors significantly associated with DS use were meeting the requirements for fruit and vegetable intake (odds ratio, OR = 1.36, 95% CI 1.02 – 1.82, p = 0.039), a belief in the importance of supplements to prevent cancer recurrence (OR = 3.13, 95% CI 2.35 – 4.18, p < 0.001) and the absence of obesity (OR = 0.58, 95% CI 0.38 – 0.87, p = 0.010).

The authors concluded that DS use among cancer survivors was common and influenced by patient’s beliefs about recurrence. They added that further work was required to better understand the reasons for such beliefs and how best to provide appropriate supplement advice to those living with a cancer diagnosis.

Citation

Conway RE et al. Dietary supplement use by individuals living with and beyond breast, prostate, and colorectal cancer: A cross‐sectional survey Cancer 2021.

Racial and ethnic disparities revealed in study of prostate MRI in men with high PSA

15th November 2021

Racial and ethnic disparities were found in a study of prostate MRI imaging among men found to have an elevated PSA result.

The presence of racial and ethnic disparities over an 8-year period have been revealed in a study of men with an elevated prostate-specific antigen (PSA) result according to researchers from the School of Economics, Georgia Institute of Technology, Atlanta, USA. Ethnic disparities in the diagnosis and treatment of prostate cancer are not new and have been associated with a complex interaction of several factors including socioeconomic status, detection at advanced stages, biological aggressiveness, family history, and genetic susceptibility. Moreover, this disparity also appears to be present among those deemed at low risk and for whom active surveillance has been advised. In fact, there is a significantly difference in prostate cancer mortality between Black and White males which is likely due in part, to low levels of PSA testing among Black, low income males.

An accurate prostate cancer diagnosis might help to reduce ethnic disparities and recently, a study using prostate magnetic resonance (MRI) has shown that this imaging modality might allow 27% of patients to avoid a primary biopsy and the diagnosis of 5% fewer clinically insignificant cancers. Furthermore, research suggests that prostate MRI is able to successfully detect prostate cancer to a similar extent in both Black and White males.

For the present study, the researchers turned to the Optum claims database which covers a diverse population and individuals from over 50 US states. They collected data from 2011 to 2017 and focused on men aged 40 years of age and older who had a single documented PSA result and no previous PSA screening or prostate MRI claims. Using PSA thresholds of above 2.5 ng/ml, 4 ng/ml and 10 ng.ml, the team set their main outcome of interest was the association between an elevated PSA result and a follow-up prostate MRI and stratified their analysis by race, ethnicity and age.

Findings

From a total of 795,809 participants with a mean age of 59.8 years, 51,500 (6.5%) had a PSA level above 4 ng/ml, of whom only 1524 (3%) underwent a subsequent prostate MRI within 180 days. When considering ethnicity, 9.6% of patients were Black, 13.6% Hispanic, 3.9% Asian and 57.3% White.

The study revealed important racial and ethnic disparities. For example, when compared to White males, Black males with a PSA of 4 ng/ml were 22% less likely to undergo a prostate MRI (odds ratio, OR = 0.78, 95% CI 0.65 – 0.89). Such ethnic disparities were also apparent for other races such that Asians with a PSA of 4 ng/ml were 24% less likely to undergo a prostate MRI (OR = 0.76, 95% CI 0.59 – 0.99). This ethnic disparity was also apparent across age groups, with Black patients aged between 65 and 74 and a PSA above 4, 23% less likely to have a prostate MRI (OR = 0.76).

The authors concluded that racial and ethnic disparities were apparent among men with an elevated PSA result in their subsequent use of a prostrate MRI. They called for future research to better understand and mitigate physician’s decision-making biases.

Citation

Abasgidze N et al. Racial and Ethnic Disparities in the Use of Prostate Magnetic Resonance Imaging Following an Elevated Prostate-Specific Antigen Test. JAMA Netw Open 2021

Kidney transplant patients with prostate cancer have similar survival

25th October 2021

Kidney transplant patients with prostate cancer should not be excluded because it does not negatively impact on overall survival.

Kidney transplant patients with prostate cancer still achieve the same benefits in terms of overall survival as those without the cancer, according to the results of a retrospective analysis by researchers from the Division of Nephrology, University Hospitals, Ohio, US. End stage kidney disease (ESKD) which requires maintenance with dialysis or a kidney transplant for survival is associated with an increased risk of death. For example, one retrospective analysis of 242 patients with ESKD, observed an annual mortality rate of 7.4%. Nevertheless, other data has found that the mortality risk from ESKD has actually decreased over the last 15 years although some work indicates how ESKD have higher levels of prostate cancer.

Kidney transplant patients have an improved quality of life compared to those continuing with dialysis but there is uncertainty over whether or not the presence of prostate cancer impacts on mortality risk in those either with ESKD or after transplantation. These were the questions to which the Ohio researchers sought answers in the present study. They turned to the US renal data system and included men aged 40 to 79 years of age with ESKD and took the main exposure to be incident prostate cancer which developed after the initiation of dialysis but before a kidney transplant. Since the clinical characteristics of patients with ESKD and prostate cancer might differ to those without prostate cancer, the researchers used propensity matching with a control group of ESKD patients but without prostate cancer and set the outcomes of interest as the time to kidney transplant and death.

Findings

There were 15,554 patients with ESKD and prostate cancer who were matched with controls, the majority of whom (47%) were aged 70 to 79 years with 42% aged 55 to 69 years. Within the matched cohort, 77.6% of patients with prostate cancer died compared to 77.1% of control patients during a mean follow-up of 3.1 years for those with prostate cancer and 3.5 years for controls. The presence of prostate cancer was associated with a 22% lower likelihood of having a kidney transplant (Hazard ratio, HR = 0.78, 95% CI 0.72 – 0.85) and an 11% higher mortality risk (HR = 1.11, 95% CI 1.08 – 1.14) compared to controls.

However when considering kidney transplant patients and using patients without a prostate cancer and no transplant as the reference point, the hazard ratios for the time to death were 0.20 (95% CI 0.18 – 0.21) for transplant patients with prostate cancer and also 0.20 for transplant patients without prostate cancer.

The authors concluded that the presence of prostate cancer in those with ESDK was associated with only a modest increased risk of death but that once these patients had a kidney transplant, the survival benefits were identical to those without cancer. The suggested the these findings indicate that in kidney transplant patients, the presence of prostate cancer should be a barrier to provision of a new kidney.

Citation

Sarabu N et al. Prostate Cancer, Kidney Transplant Wait Time, and Mortality in Maintenance Dialysis Patients: A Cohort Study Using Linked United States Renal Data System Data. Kidney Medicine 2021

FDA approves AI software to aid detection of prostate cancer

28th September 2021

AI software designed to identify an area on a prostate biopsy image with a high likelihood of cancer has received FDA approval.

Prostate cancer is the second most common cancer in men, with 1.3 million new cases recorded in 2018. Confirmation of a prostate cancer diagnosis can only be achieved via biopsy and subsequent examination of digitalised slides of the biopsy. Now, the first artificial intelligence (AI) software for in vitro diagnostic detection cancer in prostate biopsies has been approved by the FDA in the US. The software is designed to identify an area of interest on the prostate biopsy image with the highest likelihood of harbouring cancer. This alerts the pathologist if the area of concern has not been noticed on their initial review and thus can assist them in their overall assessment of the biopsy slides.

The AI system approved is Paige Prostate and it is anticipated to increase the number of identified prostate biopsy samples with cancerous tissue and ultimately save lives. The FDA approval was based on a study of Paige Prostate undertaken with three pathologists. In the study, which was conducted in two phases, each pathologist was required to assess 232 anonymised whole slide images and asked to dichotomise these as either cancerous or benign, with only 93 slides (40%) that were in fact cancerous. In the first phase, the pathologists assessed the scans alone, whereas in the second phase, 4-weeks later, the same scans were reviewed but this time using the AI software, Paige Prostate.

Findings

In the study, the Paige Prostate software alone, had a sensitivity for detecting cancer of 96% and a specificity of 98%. Without the use of Paige Prostate, the pathologists averaged a sensitivity of 74% but with the addition of the AI software, their average sensitivity increased significantly to 90% (p < 0.001). Addition of Paige Prostate mainly improved pathologists’ detection of grade 1 to 3 cancers. However, despite a greater sensitivity from the use of Paige Prostate, there was no significant difference in specificity (p = 0.327) since this was already high at an average of 97% without Paige Prostate.

Source. FDA Press release September 2021

Study shows physicians’ reluctance to use machine-learning for prostate cancer treatment planning

15th June 2021

A study shows that a machine-learning generated treatment plan for patients with prostate cancer, while accurate, was less likely to be used by physicians in practice.

Advancements in machine-learning (ML) algorithms in medicine have demonstrated that such systems can be as accurate as humans. However, few systems have been used in routine clinical practice and often ML systems tested in parallel with physicians and actions suggested by the system not acted upon in practice. To fully utilise ML systems in routine clinical care requires a shift from its current adjunctive support role, to being considered as the primary option. In trying to assess the real-world value of an ML algorithm, a team from the Princess Margaret Cancer Centre, Ontario, Canada, decided to explore the value of ML-generated curative-intent radiation therapy (RT) treatment planning for patients with prostate cancer. The team’s overall aim was to evaluate the integration of the ML system as a standard of care and undertook a two-stage study comprising an initial feasibility to clinical deployment. For the initial validation phase, the team included data from 50 patients to assess the ML performance retrospectively. The researchers delivered ML-generated RT plans and asked reviewers to assess these plans (in a blinded fashion) with the actual plans used for the patient. In the subsequent deployment phase, again with 50 patients, both physician generated and ML generated were prospectively compared, again with the treating physician blinded to the source of the plan.

Findings
The ML system proved to be much faster at generating plans than the equivalent human-driven process (median 47 vs 118 hours, p < 0.01). Overall, ML-generated plans were deemed to be clinically acceptable for treatment in 89% of cases across both the validation and deployment phase (92% duration the validation phase and 86% during the deployment phase). In only 10 cases, the ML-generated method was deemed not applicable because the plans required consultation with the treating physician, thus unblinding the review process. In addition, 72% of ML-generated RT plans were selected over human-generated RT plans in a head-to-head comparison. However, when compared to the simulation and the deployment phase, the proportion of ML-generated plans used by the treating physician actually reduced from 83% to 61% (p = 0.02).

The authors were unable to fully account for these differences and suggested that either retrospective or simulated studies cannot fully recapitulate the factors influencing clinical-decision-making when patient care is at stake and concluded that further prospective deployment studies are required to validate the impact of ML in real-world clinical settings to fully quantify the value of such methods.

Citation
McIntosh C et al. Clinical integration of machine learning for curative-intent radiation treatment of patients with prostate cancer. Nat Med 2021