Prostate cancer diagnostics : complications and ways to reduce unnecessary biopsies

Abstract

Introduction: Prostate cancer is the most common form of cancer among men in Sweden. There has been a rapid increase in the incidence rate of prostate cancer following the introduction of PSA testing and, today, more than 1800 men are diagnosed with the disease annually in Stockholm, Sweden. This increased testing has, however, not led to any significant reduction in the mortality of prostate cancer. There is no official screening programme for prostate cancer in Sweden, however, more than 60% of men above the age of 60 have undergone a PSA test in the last 5 years. What is less known is what proportion of men undergo a prostate biopsy after a PSA test and within what time frame. The majority of men undergoing a prostate biopsy are not diagnosed with a prostate cancer. In a setting where the PSA test had a better specificity these men would not have to undergo a prostate biopsy. To perform a prostate biopsy is not without risks. Serious infectious complications following prostate biopsies have been reported to be increasing in other parts of the world. The serious infectious complication rate in Stockholm, following a prostate biopsy, is not known.

Aims: To investigate if genetic markers, SNPs, can be used as a complement to PSA to predict which men with a PSA <10 ng/mL need to undergo prostate biopsies. To explore the prostate biopsy rates and results in Stockholm and to investigate when PSA testing leads to prostate biopsies and to what extent these prostate biopsies cause side effects in terms of severe infections.

Material and Methods: In Study I, 8088 men were identified who underwent at least one prostate biopsy in Stockholm between 2005 and 2007. Those alive and younger than 80 years of age, were invited to donate blood and fill out a questionnaire. 2542 men were included in the analysis when restricted to age less than 80, alive at time of invitation, valid PIN, and a PSA <10 ng/mL. In Study II, 860 men aged 50 to 69 years with a PSA of 1-3 ng/mL without a history of prostate cancer or previous prostate biopsies were invited to undergo a prostate biopsy. 172 men were stratified into low-, intermediate- and high-risk groups according to their genetic score and then underwent a prostate biopsy. In Studies I and II, a genetic score, based on the known SNPs associated with a risk of prostate cancer at the time of the study in combination with PSA and other predictive factors, was created and used in a prediction model to enhance specificity in men with a PSA<10 ng/mL and sensitivity in men with a PSA of 1-3 ng/mL. In Study III, men who had undergone at least one prostate biopsy in Stockholm from 2003 to 2012 were included. Biopsies done in 2003 were acknowledged but not included in the analysis. Migration data was used for population analysis. Data from 38 800 biopsies was analysed. Main outcome in the study was time from PSA test to prostate biopsy. In Study IV, prostate biopsies (n=44 047) done from 2003 to 2012 were included and linked by the use of PIN to microbiological data resources to identify blood cultures taken and available biograms. The main variable studied for outcome was year of biopsy. Logistic regression and time to event were used to address associations. The net reclassification index was used to evaluate the predictive performance of the genetic risk score. In all the studies men were linked to several health registers, such as the Swedish Cancer Register, the National Prostate Cancer Register, the Swedish Cause of Death Register, the National Patient Register, and the Total Population Register.

Results: In Study I, up to 23% of the prostate biopsies could have been avoided by using a genetic risk score in combination with age, family history, PSA and f/t PSA. The proportion of missed cancers would be between 5.8 and 12% depending on the risk cut-off used. The proportion of aggressive cancers missed would be between 3.3 and 8.3%. In Study II, the proportion of cancers diagnosed in the low-, intermediate- and high-risk groups was 18, 28 and 37 %, respectively (p<0.05). A borderline significant trend was seen between a higher genetic risk score and the risk of an aggressive prostate cancer. In Study III, 58 and 45% of men in aged 50-59 and 60-69 years of age, respectively, with a PSA between 4 and 10 ng/mL underwent a prostate biopsy within one year of the PSA test. For men with a PSA >10 ng/mL the proportion was 67 and 58% respectively. One out of eight men with an advanced prostate cancer had a first known PSA of >4 ng/mL more than 6 months prior to their diagnosis. In Study IV, the proportion of men with a positive blood culture within 30 days of the prostate biopsy in 2003 was 0.38 and 1.14% in 2012. Year of biopsy was highly significant as a risk factor for undergoing a blood culture and was robust both in the simple - and the adjusted analysis. Young age and low PSA values were associated with a risk of undergoing a blood culture. Men with a high Charlson Comorbidity Index had an increased risk of undergoing a blood culture. Bacteria resistant to common prophylactic antibiotics were more frequently found in blood cultures in the later years of the study than in the early years.

Conclusion: A genetic risk score can be used to enhance the sensitivity and specificity of PSA in men undergoing an investigation for prostate cancer. By reducing the number of unnecessary biopsies the number of men suffering from severe infectious complications will be reduced as well as the number diagnosed with a low-risk prostate cancer. The proportion of relatively young men not undergoing a prostate biopsy within one year of the PSA test, although their result was pathological, was surprisingly high. One way to solve this problem would be to introduce a structured follow-up after PSA testing.