Advice

Clinical and technical evidence

Clinical and technical evidence

A literature search was carried out for this briefing in accordance with the interim process and methods statement for medtech innovation briefings. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.

Published evidence

There are 7 studies summarised in this briefing including 460,503 people having prostate cancer screening in primary care.

The clinical evidence and its strengths and limitations is summarised in the overall assessment of the evidence.

Overall assessment of the evidence

Overall, the quantity of evidence for the performance of Stockholm3 in prostate cancer screening is adequate and of good methodological quality. All studies summarised here were peer-reviewed. None of the studies are based in the UK and performance may vary across different populations owing to diverse clinical practice. More prospective comparative studies are needed to evaluate the performance of Stockholm3 within the NHS setting.

Nordström et al. (2021)

Intervention and comparator(s)

Stockholm3 plus MRI-targeted biopsy was the main intervention compared with prostate-specific antigen (PSA) test and systematic prostate biopsy. Two analyses were used in the study:

  • Stockholm3 (using scores of 0.11 and 0.15 as cut-offs) versus PSA in the experimental group to enable assessment of performance when an MRI-based strategy is used for cancer detection (paired analyses)

  • PSA plus standard biopsy versus Stockholm3 plus MRI-targeted and systematic biopsy (unpaired, randomised analyses).

Key outcomes

The area under the receiver-operating characteristic curve for detection of clinically significant prostate cancer was higher for Stockholm3 (0.76; 95% confidence interval [CI] 0.72 to 0.80) compared with PSA (0.60; 95% CI 0.54 to 0.65). In the experimental group, a Stockholm3 of 0.11 or higher was non-inferior to a PSA level of 3 nanograms per ml or higher for detection of clinically significant prostate cancer (227 versus 192; relative proportion [RP] 1.18 [95% CI 1.09 to 1.28], p<0.0001 for non-inferiority), detected a similar number of low-grade prostate cancers (50 versus 41; 1.22 [95% CI: 0.96 to 1.55], p=0.053 for superiority), and was associated with more MRIs and biopsies.

Compared with a PSA level of 3 nanograms per ml or higher, a Stockholm3 of 0.15 or higher provided identical sensitivity to detect clinically significant cancer and led to fewer MRI procedures (545 versus 846; 0.64 [95% CI: 0.55 to 0.82]) and fewer biopsy procedures (311 versus 338; 0.92 [95% CI: 0.86 to 1.03]). Compared with screening using PSA and systematic biopsies, a Stockholm3 of 0.11 or higher combined with MRI-targeted and systematic biopsies was associated with higher detection of clinically significant cancers (227 [3.0%] people tested versus 106 [2.1%] people tested; RP 1.44 [95% CI 1.15 to 1.81]), lower detection of low-grade cancers (50 [0.7%] versus 73 [1.4%]; 0.46 [95% CI: 0.32 to 0.66]), and led to fewer biopsy procedures. People randomly assigned to the experimental group had a lower incidence of being prescribed antibiotics for infection (25 [1.8%] of 1,372 versus 41 [4.4%] of 921; p=0.0002) and a lower incidence of admission to hospital (16 [1.2%] versus 31 [3.4%]; p=0.0003) than those in the standard group.

This study concluded that the Stockholm3 with MRI-targeted biopsy approach for prostate cancer screening decreases over-detection without losing the ability to detect clinically significant cancer.

Strengths and limitations

Randomisation and appropriate choice of the comparator accounted for some of the strengths of this study. Another strength was the large number of people enrolled in the study. One of the limitations was non-blinding of the urologists, clinicians and urologists who do biopsies. However, the block randomisation of people included minimised allocation bias. The study was done outside the NHS. This limits the applicability of these findings in the NHS.

Karlsson et al. (2021)

Intervention and comparator(s)

Stockholm3 was the primary intervention. Stockholm3 together with 3 PSA level cut-offs was compared with no screening, and with Quadrennial screening with PSA test alone. The 3 Stockholm3/PSA combinations were:

  • PSA test and reflex Stockholm3 test for PSA levels of between 1 and 1.5 nanograms per ml

  • PSA test and reflex Stockholm3 test for PSA levels of between 1.5 and 2 nanograms per ml

  • PSA test and reflex Stockholm3 test for PSA levels of at least 2 nanograms per ml.

Key outcomes

At a PSA level threshold of 2 nanograms per ml, Stockholm3 was more effective than PSA test alone, reduced lifetime biopsies by 30%, and increased societal costs by 0.4%. Relative to the PSA test alone, the Stockholm3 with reflex thresholds of 1, 1.5 and 2 nanograms per ml, PSA levels had incremental cost-effectiveness ratio of 170,000, 60,000 and 6,000 € per quality-adjusted life year, respectively. The technology was cost effective.

Strengths and limitations

Using a lifetime horizon was 1 of the strengths of the study. The study is based on the Swedish healthcare system which limits translation of findings to the NHS setting.

Viste et al. (2020)

Intervention and comparator

Stockholm3 compared with PSA testing

Key outcomes

After 12 months of introducing Stockholm3 in Stavanger, 97% (94 out of 97) of GP clinics did prostate cancer screening using Stockholm3. Out of the 4,787 people tested, 995 (20.8%) had a positive Stockholm3 risk score (Stockholm3 risk score=11% or more), while 1,387 (29.0%) had a positive PSA test result (PSA of 3 nanograms per ml or more). There was a 28% relative decrease in the number of tested people referred for further workup. Up to 520 out of 4,784 (11%) people who tested had a positive PSA but negative Stockholm3, and 128 out of 4,784 (3%) had negative PSA but positive Stockholm3 test. The proportion of biopsies positive for cancer that showed clinically significant prostate cancer increased from 42.1% (98/223) before implementation to 64.9% (185/285) after implementation of Stockholm3 in the Stavanger region. Correspondingly, both the number and the rate of clinically non-significant cancer decreased from 135 (57.9%) before implementation to 100 (35.1%) after implementation of Stockholm3. The cost saving of implementing Stockholm3 was estimated to be between 23% and 28% because of reduced number of unnecessary MRIs, sepsis and biopsies.

Strengths and limitations

One of the strengths of this study is the choice of outcomes which are relevant to the UK. This study showed a high acceptability among GPs in Norway. There is potential for differences in clinical practice between Norway and the UK. This potentially limits generalisation of these findings to the NHS.

Grönberg et al. (2018)

Intervention and comparator(s)

Combined Stockholm3 and MRI compared with MRI alone and systematic biopsy.

Key outcomes

The study showed that Stockholm3 reduced biopsies, decreased detection of Gleason grade 1 tumours, and maintained the detection of Gleason grade 2 or more tumours. Stockholm3 combined with MRI and systematic biopsy had an acceptable sensitivity (0.94; 95% CI 0.90 to 0.97) in detecting clinically significant prostate cancer with a Gleason grade score of 2 or more when compared with systematic biopsy in all people. Stockholm3 also reduced detection of Gleason grade 1 tumours by 30% and saved 38% of biopsies from being done. When Stockholm3 was combined with MRI or targeted biopsy and systematic biopsy, it improved detection of clinically significant prostate cancer by 10% compared with systematic biopsy alone. The combined strategy of only doing MRI or targeted biopsy in people with positive Stockholm3 showed similar sensitivity to detecting clinically significant prostate cancer compared with systematic biopsy alone, but decreased detection of clinically non-significant prostate with Gleason grade 1 score. Negative predictive value for Stockholm3 was 99% when both systematic and targeted biopsy were negative.

Strengths and limitations

This study explored different clinical scenarios of using Stockholm3 and the appropriate choice of outcomes which are also relevant to the UK. One limitation is that the study was not from the UK.

Bergman et al. (2018)

Intervention and comparator(s)

Stockholm3 and PSA

Key outcomes

Biopsy was recommended in 62% of people who were referred for MRI after a positive Stockholm3 test. Of those having a biopsy, 58% had high grade cancer while only 6% had low-grade cancer. A health economic analysis reported Stockholm3 with MRI followed by targeted and systematic biopsies had the lowest costs when compared with 1) PSA then systemic biopsy if PSA level is elevated (PSA level of more than 3 nanograms per ml), 2) PSA then MRI if PSA level is raised, then systemic biopsy if positive MRI results. The study showed the effectiveness of using a reflex testing model. Using nurses in the screening reduces visits to the urologists for patients that do not need a biopsy. The cost savings of implementing Stockholm3 was estimated to 17% because of reduced number of unnecessary MRIs, biopsies, and treatments.

Strengths and limitations

Some of the strengths for the study included study design, comparison of 3 strategies and including up to 10 primary care sites. One of the limitations is that the study was done outside the NHS. The study did not provide study design details.

Ström et al. (2017)

Intervention and comparator

Stockholm3 compared with PSA testing.

Key outcomes

The study looked at updating the Stockholm3 algorithm to improve its performance in prostate cancer diagnosis. When used as a reflex test for people with PSA levels of at least 3 nanograms per ml, Stockholm3 reduced the number of biopsies needed by 34% compared with using PSA levels alone, with equal sensitivity.

Strengths and limitations

One limitation was that the population was ethnically homogenous. This limits generalisation of the performance in other ethnic groups.

Grönberg et al. (2015)

Intervention and comparator

Stockholm3 compared with PSA testing.

Key outcomes

The Stockholm3 model performed better than PSA testing alone in detecting high grade cancers with a Gleason score of at least 7. Stockholm3 was also reported to reduce the number of biopsies by 32% and avoided up to 44% of benign biopsies.

Strengths and limitations

One limitation was that the population was ethnically homogenous. This limits generalisation of the performance in other ethnic groups.

Sustainability

The company claims the Stockholm3 may help reduce the environmental impact by decreasing energy use and travel. There is no published evidence to support these claims.

Recent and ongoing studies