Scientists identify blood proteins linked to aggressive prostate cancer years before diagnosis, raising hopes for better risk prediction while uncovering potential roles for immune activity and cancer-related pathways.
Study: Circulating Immune and Cancer Driver Gene Biomarkers Precede Prostate Cancer Diagnosis by over 10 yr. Image credit: Kitsawet Saethao/Shutterstock.com
A recent study published in the journal European Urology reports associations between some circulating plasma proteins and prostate cancer, with some predating diagnosis by over a decade.
Protein markers of aggressive prostate cancer
Prostate cancer is the most commonly diagnosed cancer among men, but risk prediction remains difficult because little is known about its risk factors or etiology. Prior research has identified blood proteins associated with prostate cancer risk, such as IGF-1, MSMB, and immunosurveillance proteins like FLT3LG and IL-15.
Using high-throughput proteomics, such associations can now be studied on a large scale. The current study aimed to identify circulating proteins prospectively linked to prostate cancer, and specifically to aggressive disease.
Researchers analyze over 6,000 proteins in European men
The investigators drew on data from the European Prospective Investigation into Cancer and Nutrition (EPIC), a European cohort of over 500,000 people recruited in 1992-2000. This study used men from Italy, the Netherlands, Spain and the UK.
It included 982 men who developed prostate cancer, alongside a randomly selected subcohort of 1,573 men, including 79 who developed the disease. The median potential follow-up was 16.5 years.
The researchers measured 6,412 plasma proteins (7,363 aptamers). The primary endpoints were overall and aggressive prostate cancer (982 and 290 cases, respectively). Aggressive tumors were defined as high grade, advanced stage, antigen">PSA above 20 ng/mL at diagnosis, or death from prostate cancer.
They looked for associations between each of the circulating proteins and prostate cancer risk using Prentice-weighted Cox regression analysis, stratified by age at recruitment and by center. The analyses were adjusted for body mass index, smoking, alcohol, and education. Family history and ancestry were not available. The researchers also applied correction for multiple testing within each endpoint.
Baseline characteristics reveal differences in prostate cancer cases
Men who developed prostate cancer were generally older when they joined the study, with a median age of 57.2 years, compared with 50.6 years among men in the randomly selected comparison group. They were also less likely to smoke or consume large amounts of alcohol.
Three blood proteins linked to overall prostate cancer risk
The study showed 6 aptamers corresponding to 3 proteins associated with the overall risk of prostate cancer after adjustment and correction for multiple testing. These included KLK3 (the PSA gene), by far the strongest, associated with about 76% to 128% higher hazard per standard deviation (SD) increase; ACP3 (prostatic acid phosphatase) with 20% elevated hazard; and FLT4 with 26% higher hazard.
Associations involving KLK3, which encodes prostate-specific antigen (PSA), and ACP3, which encodes prostatic acid phosphatase, were expected because both proteins are associated with prostate tissue. FLT4 (a VEGF-C receptor involved in lymphangiogenesis) was associated with almost all the endpoints. Previous experimental studies suggest that the VEGF-C/FLT4 pathway may contribute to tumor invasion.
Seven proteins associated with aggressive prostate cancer risk
Seven proteins were associated with aggressive prostate cancer, including ANKRD1 (HR 1.25) and TPM3 (HR 1.25). ANKRD1 was significantly associated (23% to 29% increase in hazard) across all three categories of aggressive disease: high-grade, advanced-stage, and lethal disease.
Blood protein differences emerge over 15 years before diagnosis
The timing of prostate cancer diagnosis revealed further differences in blood protein levels. Nine proteins were associated with cancers diagnosed within five years of blood collection, potentially reflecting the presence of undiagnosed disease. Another 12 proteins were linked to cancers diagnosed more than 15 years later, suggesting that some associations may emerge long before diagnosis.
Among these, higher levels of GZMK and ANXA2 were associated with a 43% and 38% lower hazard of prostate cancer, respectively. Immune cells that help destroy abnormal cells produce GZMK, raising the possibility that higher blood levels reflect stronger immune surveillance. However, the researchers suggest that lower circulating levels could also indicate that these immune cells have moved from the bloodstream into prostate tissue. The association between higher GZMK levels and lower prostate cancer risk became stronger as the time between blood collection and diagnosis increased.
In contrast, inflammatory proteins such as IL-17B and OSCAR were associated with a higher risk of prostate cancer over the longer term.
The researchers also identified proteins associated with different forms of aggressive prostate cancer, including 12 linked to high-grade tumors, nine to advanced-stage disease, and 14 to lethal prostate cancer. For every standard deviation increase in protein levels, TFPI2 was associated with a 39% higher hazard of high-grade cancer, AMIGO2 with a 49% higher hazard of advanced-stage disease, and HLA-DRB3 with a 39% higher hazard of lethal prostate cancer.
Overall, 53 proteins were associated with at least one prostate cancer outcome, highlighting a range of potential biological markers linked to disease development and progression.
Cancer driver-gene proteins linked to future aggressive disease
The researchers found that of 71 driver genes tested, circulating proteins encoded by ERBB3, KRAS and NRAS were associated with aggressive disease. RBFOX1 and CTNNB1 were associated with prostate cancer diagnosed more than 15 years and 5-10 years after blood draw, respectively.
Independent studies provide mixed support for protein associations
The researchers compared their findings with other studies, but the results were mixed. In the ARIC study, 28 proteins were available for comparison, with KLK3 and ACP3 showing associations with several prostate cancer outcomes. ANXA2 and APOC2 showed similar patterns, while findings for OSCAR, AMIGO2 and FLT4 varied by ancestry.
In the UK Biobank, none of the 18 proteins available for comparison showed significant associations with prostate cancer, potentially reflecting differences between the protein measurement methods used.
Genetic analyses provided limited evidence that the proteins play a causal role in prostate cancer. Although four proteins showed associations consistent with the original findings, further genetic evidence did not support firm causal conclusions. A separate analysis identified eight genes associated with prostate cancer risk, including ACP3.
The researchers suggested that differences in study populations, cancer outcome definitions, and protein measurement methods may help explain the inconsistent findings.
Missing tumor data complicates aggressive cancer risk estimates
The study has several limitations. Its observational design cannot establish whether the proteins contribute to prostate cancer development, while missing tumor data and inconsistent findings across other studies limit confidence in some associations. Further research is needed to determine whether these proteins can improve prostate cancer risk prediction beyond existing clinical markers such as PSA.
Protein biomarkers require further validation before clinical use
This large discovery study generates hypotheses about various proteins involved in inflammation, immune surveillance, adhesion, and angiogenesis in prostate cancer, with ANKRD1 being a candidate marker of aggressive disease. However, these exploratory results do not demonstrate that these proteins improve risk prediction beyond PSA and standard factors. Dedicated prediction studies will be required to assess their clinical utility for risk stratification.
Study identifies blood proteins associated with aggressive prostate cancer risk