4. Early detection – Prostate Cancer

Curative treatment of prostate cancer is only possible in the organ-confined stage. Prostate cancer usually has a slow “natural” progression if untreated, and therefore only men with a life expectancy of more than 10-15 years benefit from curative treatment. The goal of early detection is to identify organ-confined aggressive tumors in asymptomatic men with a presumed life expectancy of at least ten years. Earlier detection will also detect cancers that would never have become symptomatic without early detection or screening. Thus, the search for prostate carcinomas that are still curable also leads to unnecessary diagnosis and treatment with the associated side effects.

Previous randomized PSA-based screening studies of prostate carcinomas can be considered heterogeneous and there is no broad consensus on how they should be interpreted. They refer to a PSA value determined at different age cut-offs and intervals as part of population-based screening. At a certain cutoff value, which was set at different levels in the various studies, further diagnostic procedures are then derived from this PSA value. In addition to the difficulty in interpretation, the treatment approach in Germany is generally different. On the one hand, there is early detection in the sense of opportunistic screening if desired and there is an opportunity for this. On the other hand, it is quite common to initially monitor a borderline PSA value over time and to determine an indication for a prostate punch biopsy based only on an abnormal rise in PSA. Theoretically, this leads to a significant decrease in the number of prostate-tissue biopsies required, and to a significant decrease in the number of patients unnecessarily burdened with treatment, although evidence is lacking in this regard.

Drafting of this chapter of the guideline has been complicated due to the published data not allowing for a consensus on interpretation, on the one hand, and the issue of transferability to the realities of German healthcare, on the other hand.

The benefits of prostate cancer screening

A total of five RCTs are available on the efficacy of PSA-assisted screening (ERSPC, PLCO, Québec, Stockholm, Norrköping), three of which are of poor methodological quality. In a multicenter European study (ERSPC) (n = 162,243), a significant reduction in prostate cancer-specific mortality was shown in the age group of 55 to 69 years after a median observation period of 13 years [57]. In absolute terms, for every 10,000 men screened, 12 fewer died of prostate cancer (49 vs. 61). In an intention-to-screen analysis, this corresponds to a number needed to screen (NNS) of 781 (95% CI 490–1929). The study showed no difference in all-cause mortality between the intervention and control groups, but the ERSPC was not sufficiently powered to demonstrate such a difference. The results of the study were based substantially on the contribution of the Swedish cohort (n =11,852; age 55-64). The Swedish cohort showed the highest incidence of prostate cancer among European countries. In a separate publication evaluating the entire Swedish cohort (n =19,904; age 50-64), prostate cancer-specific mortality was reduced from 0.9% in the control group to 0.5% in the screening group after 14 years [58]. In a multicenter American RCT (PLCO) (n=76,700), there is no effect on prostate cancer-specific mortality after 13 years. However, this study reports 86% contamination of the control arm [59], which means that the results are not sufficiently powered and it is likely an effect is being overlooked. Taking into account the lack of compliance in the American study (PLCO), on the one hand, and contamination, on the other hand, the percentage reduction in prostate cancer-specific mortality is shown to be comparable to that of the European study (ERSPC) at 25-30% [60][61].

Meta-analyses pooling the data from the RCTs - despite their heterogeneity - show no reduction in prostate cancer-specific mortality or overall survival in the overall group, nor for different age groups [62]. The IQWIG report on prostate cancer screening by means of PSA testing of 06/02/2020 arrived at comparable results (IQWIG report S19-01 2020). However, with regard to prostate cancer-specific mortality, there was an indication of a benefit of prostate cancer screening by means of PSA testing on the basis of the studies with a PSA cutoff value below 4 ng/ml. For other groups with a PSA cutoff value of 4 ng/ml and higher, there was no evidence of benefit.

Comparison of the different screening programs (in terms of interval, PSA cutoff, populations screened, participation in the intervention and contamination in the control group, overall disease risk, and follow-up periods) in the European study (ERSPC) shows that the magnitude of reduction in prostate cancer-specific mortality does not depend on the initial PSA cutoff < 4 ng/ml or > 4 ng/ml, but predominantly depends on participation in PSA screening, biopsy compliance, and screening frequency [63].

[57]

On the harms of prostate cancer screening

All RCTs on PSA-assisted screening show a significantly increased number of detected prostate cancers in the screening group and thus a significant rate of overdiagnosis: In the European study, 1,301 more cases of prostate cancer were diagnosed in the screening group than in the control group (7,408 vs. 6,107), corresponding to a cumulative incidence of 10.2% and 6.8%, respectively [57]. Subsequently, 34 additional prostate cancer patients had to be detected in the screening group for one death from prostate cancer to be statistically prevented.

The results after 13 years, expressed per 10,000 men:

• per 10,000 men, 1,020 prostate cancers were detected in the screening group. 49 men died of PCa
• per 10,000 men, 680 prostate cancers were detected in the control group. 61 men died of PCa
• for each PCa death prevented, there were 34 men with a detected prostate carcinoma (number needed to be detected)

In the Swedish study, 420 more carcinomas were diagnosed in the screening group than in the control group over a 14-year period (1,138 vs. 718) [58]. This corresponds to an absolute incidence increase of 4.5% from 8.2% to 12.7% (greater than 50% in relative terms compared to the control group). The “Number Needed to Screen" after 14 years is reported as 293; the number of additional diagnoses to prevent one death from prostate cancer was 12. In contrast, there are 120 more impotent men per 10,000 screened due to the additional, mostly radical surgical treatments carried out as well as 25 more incontinence sufferers per 10,000 (extrapolated data from a before-after comparison of a subgroup of operated patients) [64].

The IQWIG report from 2020 concludes that there is evidence of harm regarding the consequences of overdiagnosis as well as false-positive screening findings in a situation where overall data is scarce. It concludes that the benefits of prostate cancer screening by PSA testing do not outweigh the harms.

There is no broad consensus on how to interpret the benefits in terms of potential quality of life gained or lost. For example, Heijnsdiek et al. (2012) described the quality of life gained (QALYs) in particular considering all the steps of diagnostic screening (overdiagnosis), therapies (overtreatment) and side effects [65]. This showed that per 1,000 invited participants ultimately 73 life-years (QALYs = quality-adjusted life years; years in “complete“ health) resulted. After deduction of the damage side of the equation due to overdiagnosis and overtreatment, a gain of 56 years of life (QALYs) remains [65]. Further optimization of the screening procedure to reduce harm seems possible by providing surgery in “high volume“ centers, stopping screening in older men, and even more frequent use of Active Surveillance. With this strategy, a gain of 74 (+18) life-years (QALYs) could be achieved (Carlson et al; Cancer, 2016). Relating the number necessary to invite (NNI) and number necessary to diagnose (NND) to the remaining lifetime of patients in the Miscan model (NEJM) results in an overall NNI and NND of only 98 and 4, respectively, to save one life for the European study (ERSPC). For the Gothenburg study, these values were only 53 and 2 respectively [65].

However, IQWIG does not recognize any clear or significant results on comparative health-related quality of life in the publications by Heijnsdijk from 2012 and Carlsson from 2016, as both publications performed microsimulations based on ERSPC data. IQWIG considers these models to be subject to large uncertainties, as they inevitably contain assumptions regarding all aspects for which robust evidence is lacking, which may or may not be true.

Adding to this ambiguity in the data, there are suggestions from more recent data that PSA may be a prognostic marker for the long-term evolution of prostate cancer, thereby reducing NNI and its follow-up. Men in the top range of the PSA percentile (>2.4ng/ml) aged 51-55 years contribute to over 50% of long-term prostate cancer deaths, whereas men with a PSA <1.0 ng/ml have a 0.9% chance of dying from prostate cancer [66]. These retrospective data could also be proven prospectively in the European study (ERSPC). Men with PSA <1.0ng/ml (50% of the total group) were virtually cancer free for 12 years [67]. This strategy could possibly be used to define small risk groups requiring more intensive surveillance early on, while allowing the majority of men to be screened at extended follow-up intervals (4-6 years) [67][68]. However, the evidence of a clear benefit due to this strategy in the early detection of prostate cancer with an advantage for a screening group that can be proven has not been provided so far.

Re: Recommendation 4.1

The recommendation to provide information corresponds to good clinical practice and the physician's duty to inform. It is similarly found in the guidelines of the European Association of Urology (EAU) [69]. Currently, education on the early detection of prostate carcinoma should be provided on one occasion.

The statutory offer of screening in Germany (for men from the age of 45 years, provided once per year) currently includes taking a history of any complaints or other health changes, the examination of the external genitalia, and palpation of the prostate and lymph nodes.

There is a consensus among all the professional societies and associations involved in this chapter that information should be provided opportunistically on the early detection of prostate cancer and open, non-biased advice given on the associated advantages and disadvantages.

The German Association for Prostate Cancer Self-Help (BPS), the Working Group for Prevention and Integrative Medicine in Oncology (PRiO), the German Radiological Society (DRG), the Professional Association of German Radiotherapists (BVDST) and the German Society for Urology (DGU) recommend that all men who are at least 45 years old and have a presumed life expectancy of more than 10 years be informed about their options for early detection; the opportunity to do this has deliberately not been narrowly defined, as it can and should vary depending on the situation and range from advice in the context of treating another disease to a patient expressing a desire for more information about their general health.

According to the German Society for General Medicine and Family Medicine (DEGAM), this occasion can and should be situationally diverse and range from consultation in the context of the treatment of another disease to the desire for general health information and also only exist when men raise the topic of early detection of prostate cancer of their own accord. Thus, men can (but are not required to) be proactively approached about early detection of prostate cancer.

Men interested in early detection should be made aware of the risk of overdiagnosis and overtreatment in absolute terms, as outlined above. Based on the most recent data from the ERSPC study [57], the following figures can be used for clarification:

By providing comprehensive information, they should be able to decide for themselves whether they wish to undergo screening procedures. The information should be provided in accordance with the individual's need for information and understanding.

The German Guideline Program in Oncology offers a patient guideline on screening for educational purposes that can be downloaded free of charge: http://www.leitlinienprogramm-onkologie.de/patientenleitlinien/prostatakrebs/.

An electronic decision aid such as arriba PSA (https://arriba-hausarzt.de/module/psa-screening) is recommended, which includes information on the significance of positive and negative test results, or https://www.krebshilfe.de/fileadmin/Downloads/PDFs/Praeventionsfaltblaetter_Frueherkennung/FB-428_Prostatakrebs-erkennen.pdf.

The possible consequences of a biopsy in the case of a positive PSA value and, if necessary, MRI of the prostate should also be pointed out; see the chapter on diagnostics.

The information should be provided by the attending physician and documented in case of a legal dispute due to an allegation of inadequate information being given, since if there is no documentation, the burden of proof falls to the physician who provided the information. Various auxiliary instruments (e.g. information materials available in the practice) can be used or trained assistants (e.g. physician's assistant) can be involved. The final consultation for providing information to the patient is the responsibility of the attending physician. Clarification is also possible – or can be additionally sought from independent institutions providing advice. In particular, the German Cancer Research Center (DKFZ) should be mentioned in this respect (https://www.krebsinformationsdienst.de/tumorarten/prostatakrebs/psa-test-frueherkennung.php).

Re: Recommendation 4.2 and Statement 4.3

Digital-rectal examination alone is not considered suitable for early detection. The primary procedure for early detection is determination of PSA value because the sensitivity of PSA determination is higher than the sensitivity of other procedures [47][48][55][49][50][70][46]. Imaging techniques have significantly lower specificity in addition to lower sensitivity [46][47][48][55][49][50].

The combination of PSA level plus DRE increases the specificity of prostate cancer detection [46][47]. In the systematic review by Harris et al. for screening studies, cancer detection with PSA (cutoff 4 ng/ml) is reported as 4.6 % and in combination with DRE as 5.8 % [46]. Analyses of the PLCO screening study [51][52] show that both a suspicious DRE finding and a PSA level above the defined threshold are associated with the detection of clinically significant prostate cancer; the hazard ratios are higher for PSA than for DRE in each case (5.48 vs. 2.21; 5.23 vs. 2.54). Furthermore, more clinically significant cancers were found due to abnormal PSA alone compared with abnormal DRE alone (detection of one clinically significant prostate cancer per 4.4 men biopsied vs. one per 6 men) – more biopsies were also performed due to abnormal PSA compared with suspicious DRE (32.6% vs. 21.6%). Of the men with suspicious DRE who were found to have prostate cancer on biopsy, 69% had a concurrent abnormal PSA level. Due to the weaker evidence, the authors have made a subordinated recommendation for DRE as an additional procedure.

Re: Recommendation 4.4

Testing PSA levels in the blood is not part of the statutory screening program, but many medical practices in Germany offer it as an “individual health service“ (IGeL). It is important to note that a PSA test result can only predict the risk of prostate cancer.

Any PSA value which would give rise to further, potentially invasive, diagnostic tests (biopsy) requires initial confirmatory testing if its plausibility cannot be sufficiently proven by the known trend of previous PSA measurements.

Before further diagnostic testings, factors influencing sample storage or sample transport should be excluded.

Other factors relating to the patient can influence the PSA value. For example, urinary retention, acute prostatitis, or previous manipulations of the prostate (e.g., DRE, catheterization, colonoscopy) should be considered. There are contrdictions in the literature as to whether cycling or ejaculations significantly alter PSA levels. It is likely that the total PSA value is only marginally affected by this, but the free PSA value may be somewhat more affected. In case of doubt, a confirmatory measurement after a few days, having removed the influencing factor(s) mentioned, can provide clarity. The prostate volume can also have an influence on the level of the PSA value, see accompanying text to recommendation 4.6.

PSA measurement procedure

In accordance with good clinical practice, the following points should be observed for the PSA measurement procedure:

• Quantitative determination of PSA concentration is required for early detection. Semiquantitative or qualitative strip or threshold tests do not measure the PSA ranges that are relevant for early detection examinations with sufficient precision and accuracy (see below).

• The measured values obtained are interpreted with consideration for the PSA measurement method used and its calibration, only PSA measurement methods for which reference ranges are individually defined are used. The level of PSA concentration determined by different PSA measurement methods depends on the calibration of the measurement system. Calibration with WHO International Standard 96/668 or 96/670 leads to measurement results with approx. 20 % lower PSA concentrations in identical samples compared to calibration as per Hybritech [71]. Since the measured values of different measuring systems may differ in a clinically significant manner even despite identical calibration, it is recommended that the reference range determined specifically for this method be specified for each measuring system when interpreting the measurement result. For meaningful interpretation of the measurement result, the laboratory report must therefore contain the following information:
• Name and manufacturer of the PSA measurement method;
• Indication of the calibration for the measurement system (WHO, Hybritech, others);
• The normal or reference ­range determined for this measurement method in a sufficiently large group of men of appropriate age, with an indication of the sensitivity and specificity valid for this reference range for the detection of prostate carcinoma in prostate biopsies.

In the absence of this information, it is recommended that the laboratory or the manufacturer of the PSA measurement method be consulted before interpreting the test result [72].

Lacking reliability of test strips

In addition to the conventional quantitative measurement methods for PSA, semiquantitative methods or qualitative methods are also offered, so-called test strips. The reliability of test strips investigated to date is insufficient, particularly in the PSA concentration range that is relevant for early detection. In addition, PSA trends over time cannot be detected with test strips in serial measurements [73][74][75][76].

Re: Recommendation 4.5

This recommendation includes the start time, time intervals and end of PSA-based screening. Since opportunistic rather than population-based screening is practiced in Germany, the work which the recommendations are based on is only applicable with certain provisions.

The recommendation of starting at 45 years is a compromise. The current AUA recommendations, published in 2013 and reaffirmed in 2018, do not recommend routine screening in the 40-54 age group [77]. On the other hand, the current 2020 EAU–guideline recommends “baseline“ PSA testing as early as age 40 [69].

The specific recommendation for the 45-50 age group is based on the fact that a homogeneous PSA range at baseline may help to improve the usability of PSA kinetics and to detect any dynamics early. While the recommendation deliberately refrains from differentiating by risk groups due to lack of relevant evidence, several papers have been able to demonstrate the association between progression of cancer disease and an increase in PSA levels, e.g., an analysis by Lilja et al. (2011) [56]. Based on a Swedish cohort of men who gave blood samples aged 33-50 years (n = 21,277), they compared the baseline PSA levels of 1,312 men who had developed prostate cancer after a median of 23 years with the PSA levels of a total of 3,728 healthy men (1:3 case-control study embedded in a cohort study). The results of the study show that the risk of prostate cancer among men in the 44-50-year risk group increases with the level of baseline PSA. This relationship also holds for the subgroups of palpable and advanced cancer. The absolute risks determined in the work of Lilja et al. 2011 for the respective PSA values and the screening intervals derived from them by Lilja et al. are outlined below.

The specific recommendation regarding the time intervals for a PSA determination is also based on the study results; they suggest that an annual screening examination is probably not useful in general. For example, the number of patients with initial PSA levels rising to 4.1 ng/ml, making biopsy reasonable, ranges from 0 % to 4 % [78]. A paper from the ERSPC study of 5,771 men showed only 4 % of patients with an initial PSA of 3 ng/ml within four years [79]. The reduction in mortality shown in the ERSPC study [80] is based on a screening interval of 4 years, justifying this interval of once per year for men from the age of 45 years.

This theory is supported by a 2003 paper by Van der Cruijsen-Koeter [81]. In a series of 8,350 men, 18 interval carcinomas (i.e. carcinomas missed by screening) were found within four years in those with an initial PSA below 3 ng/ml, all of which were stage T2a or less and thus could be treated with curative intent [81]. A 2005 subgroup analysis by Roobol [82] of over 1,200 men between 55 years and 65 years and an initial PSA of less than 1 ng/ml showed that only eight men in this group developed prostate cancer after eight years [82]. Roobol et al. concluded for men with an initial PSA level < 1 ng/ml a screening interval of 8 years to be sufficient. However, the intervals should not be too long due to the paradigm shift in the interpretation of PSA level trends.

Regarding the issue of an upper age limit for regular screening­, the prospective case-control study by Vickers et al. 2010 of approximately 100 men with an initial PSA level at 60 years of age showed that few developed prostate cancer within 25 years (19 %) [83]. 7 % had metastases after 25 years and one death from prostate cancer was recorded. Although it should be taken into account that life expectancy has increased dramatically in the last 25 years, in principle, no further PSA screening should be performed in men aged 70 years and older.

The formulation of this recommendation does not affect the individual decision of the physician to decide, in consultation with the patient, on a longer or shorter interval for early detection and, if desired, to perform this at younger or older ages.

Re: Recommendation 4.6

≥ 4 ng/ml, influencing factors:

The cutoff value of 4 ng/ml was described as reasonable in a large study by Catalona et al. (1994) [84] in over 6,000 subjects [71]. However, an adjustment of the cutoff value depending on calibration of the measurement method used is required. In the placebo group of the Prostate Cancer Prevention Trial, in which 2,950 men with PSA values below 4 ng/ml underwent prostate biopsy, there was no PSA threshold above which the incidence of prostate cancer increased abruptly, but rather a linear relationship between the level of the PSA value and cancer detection [85]. So far, the question remains unanswered as to which strategy can be used to reduce the number of resulting unnecessary biopsies and allow for overtreatment to be avoided.

The PSA threshold of 4 ng/ml was determined in “healthy“ men from the normal population, who may differ in screening-relevant characteristics from men in screening studies. Patients in urology practice are more likely to have micturition symptoms, residual urine, enlarged prostate, or urinary tract infections. For a meaningful interpretation of the PSA threshold, special characteristics should be taken into account in all cases except where there is an exclusive desire for a screening examination. Prostate enlargement or a urinary tract infection can lead to an increase in the PSA value without necessarily increasing the risk of prostate cancer. To avoid unnecessary prostate biopsies, transrectal ultrasound sizing of the prostate can be helpful, as a larger prostate volume can influence the PSA level. Prostate size should therefore be considered when interpreting the PSA value [86][87][88][89]. Another decision aid for or against a prostate biopsy is the determination of free PSA, its ratio increases with the increasing volume in benign prostatic hypertrophy. Men with a high ratio of free to total PSA (approx. > 0.24, with threshold value depending on the measurement method) have a lower risk of a clinically significant prostate carcinoma being present [89][90][91].

If, with a slightly elevated total PSA value, a constellation of findings is present that indicates only a low risk of prostate cancer, the option of further PSA follow-up testing should be discussed with the patient in addition to the option of a prostate biopsy.

Abnormal PSA rise:

Carter et al. observed 980 patients over nearly 10 years. They showed that when a threshold of 0.35 ng/ml per year was exceeded, cancer-specific survival decreased from 92 % to 54 %. Another study group was able to show that a PSA increase­ at a rate of 0.5 ng/ml per year in 4,800 men was the best predictor for the identification of prostate cancer [92]. In a study of 6,844 men, the ERSPC study reported a specificity for a PSA rate of rise of 0.4 ng/ml per year of 92 % [93].

However, the absolute level must be taken into account; with the recommendation of very low threshold values for the PSA rate of increase such as 0.35 ng/ml per year by the National Comprehensive Cancer Network Clinical Practice Guideline on Prostate Cancer Early Detection from 2010 [94], there is a high risk that the biological variability of the PSA value in itself could cause this threshold value to be exceeded, particularly where short observation intervals are used without there being an underlying prostate carcinoma [95].

In summary, more comprehensive investigation should be considered if the first screening examination returns a PSA value ≥ 4 ng/ml (Hybritech). Over time, the indication for biopsy can be individually determined by the trend in PSA levels, whereby the threshold value should range between 0.35 ng/ml per year and 0.75 ng/ml per year. Minimum quality assurance standards for PSA measurement should be followed. The PSA values should be checked after six to eight weeks.