Using systematic TRUS-guided biopsy at diagnosis, as historically has been the only option until the development of MRI for prostate cancer detection, inherently results in random and/or systematic sampling error [22]. Random sampling error is caused by chance; i.e. by systematicly determining the location of the biopsy cores, they are inadvertently not taken at the location of the greatest diameter or highest grade of a lesion. Systematic sampling error means that a part of the prostate is always overlooked, for example, the anterior part because it is more difficult to adequately sample. Both ways of misclassification play a role at diagnostic systematic TRUS-guided biopsies, as most often 8–14 biopsy cores are taken randomly at different locations from mainly the peripheral zone of the prostate. In contrast, MRI-guided biopsies are taken from lesions suspicious for prostate cancer on MRI and therefore are less susceptible to sampling error but are still depending on the accurate detection of suspicious lesions on MRI (see Chaps. 10 and 11).

Misclassification rates in men selected for AS can be determined by using data from radical prostatectomy series in men otherwise eligible for AS. At radical prostatectomy, 33% to 45% of men who met the study inclusion criteria of six different AS protocols appeared to have stage T3 or greater or Gleason ≥7 [18]. In another radical prostatectomy study, looking at an intensively screened population [19], the rate of misclassification for three AS studies (Johns Hopkins University, University of Toronto, and “Prostate cancer Research International: Active Surveillance” (PRIAS)) ranged from 19% to 32%. On average, about one third or more of patients are misclassified after a diagnosis by systematic TRUS-guided biopsy [23] (Fig. 8.1).

As misclassification at diagnosis with systematic TRUS-guided biopsy is substantial (about one third of patients), reevaluation of tumor stage and grade is essential [18, 19]. To confirm the tumor stage and grade at diagnosis, the majority of AS protocols include a repeat systematic TRUS-guided biopsy within the first year. Repeating a systematic TRUS-guided biopsy ensures an appropriate number of biopsy cores are taken at the right locations and decreases the possibility of (mainly random) sampling error [24]. Currently, many AS studies have also incorporated MRI in their protocols; the diagnostic accuracy and role of MRI within AS protocols will be discussed in Chaps. 10 and 11.

Reclassification rates at confirmatory or first risk assessment vary between AS cohorts depending on how strictly patients are selected at inclusion and the reclassification criteria used. Few studies report on reclassification at confirmatory or first risk assessment separately. However, the PRIAS study showed a reclassification rate to Gleason ≥3 + 4 or high-volume prostate cancer at first systematic TRUS-guided biopsy of 24%, around 1 year after enrollment [11]. The University of Toronto showed a systematic TRUS-guided biopsy reclassification rate to Gleason ≥3 + 4 of 23%, at a median of 1.4 years after enrollment [25] (Table 8.2). Unfortunately, not all misclassified patients seem to be detected by confirmatory or first risk assessments, which can be derived from the following observations: reclassification rates (23%–27%) are lower than misclassification rates (19%–45%) as shown above, and in addition, a considerable part of reclassified patients prove to have undergone a possible unnecessary radical treatment (as will be discussed further on in this chapter). However, anxious to avoid missing the so-called window of curability, all protocols include at least one (confirmation) systematic TRUS-guided biopsy to detect misclassified men (Table 8.1). Evidence is sparse for when the timing of a confirmatory risk assessment is optimal. The Memorial Sloan Kettering Cancer Center (MSKCC ) protocol performs a confirmatory systematic TRUS-guided biopsy at 3 months after diagnosis where other protocols require this between 6 and 12 months after diagnosis (Table 8.1). It is reasonable to assume, based on studies comparing delayed and immediate radical treatment, that very few men will experience unfavorable results from extending the time between diagnoses and a confirmatory biopsy procedure, as is done in current protocols (up to 1 year) [26]. Therefore, it seems responsible to involve the patient in the decision-making and perform a confirmatory risk assessment (including MRI if available) at his convenience, somewhere within the first year after diagnosis.

Next to misclassification, progression over time adds to the uncertainty of the actual tumor grade and volume during follow-up. Pathologically confirmed Gleason grade ≤ 3 + 3 prostate cancer, after radical prostatectomy, has negligible potential to metastasize [27], as discussed in Chap. 5. However, in untreated patients, new lesions may develop and an initial Gleason 3 + 3 prostate cancer can progress to Gleason 3 + 4 or higher. Whether true grade progression occurs from Gleason 3 + 3 to Gleason ≥3 + 4 has long been a topic of discussion as empirical evidence is lacking. However, two modeling studies concluded that Gleason grade progression does occur over time [20, 21]. It is estimated, by modeling the outcomes of serial biopsies during AS corrected for misclassification, that the likelihood of true grade progression from a Gleason 3 + 3 to Gleason ≥3 + 4 over a period of 10 years ranges from 12% to 24%, translating to a yearly grade progression risk of 1%–2% [21] (see Fig. 8.1). Therefore, repeated risk assessments remain essential to confirm absence of progression.

In AS protocols, several methods are used to detect progression. Some, like Gleason grade on biopsy, are a direct indication of the presence of more aggressive disease that likely needs curative treatment. Others, like PSA kinetics (PSA-DT or PSA velocity) or an increase of the number of tumor-positive biopsy cores, indirectly indicate that more aggressive disease is present. Whether these surrogate markers of disease progression should indeed trigger curative treatment depends on their direct relation with disease progression.

PSA tests can be repeatedly performed without serious drawbacks and are performed frequently during the course of AS to monitor PSA level changes. An important factor which complicates the interpretation of PSA tests is the clinical context in which it is measured, as a rise in PSA level can be related to both benign (e.g., prostatic hyperplasia or infection) and malignant origins. According to most literature, a high PSA, PSA density, and Prostate Health Index (PHI) or fast PSA kinetics may be indicative of unfavorable prostate cancer on biopsy or radical prostatectomy [7, 28], although not all studies agree on their value for clinical decision-making [29]. As discussed in [11, 29, 30], adverse PSA kinetics should not prompt radical treatment; instead (and depending on clinical context), they should be used as a trigger for stricter follow-up (e.g., more frequent PSA measurements, an MRI, and/or TRUS biopsy). PSA density thresholds from 0.10 to 0.20 ng/ml/ml are associated with abnormalities on MRI, biopsy reclassification, and unfavorable pathologic characteristics [28, 31] where the threshold of 0.15 ng/ml/ml is most frequently used. The PHI test (a combination of total PSA, %fPSA, and proPSA) is shown to improve predicting unfavorable outcomes on biopsy and radical prostatectomy [32, 33] but is currently infrequently incorporated in AS protocols. Lastly, the frequency of PSA testing most often lies between two and four times per year, and decreasing the frequency of PSA testing after 2–4 years of stable PSA values, as progression after that time is rare, is a potential option to reduce a part of the burden of long-term AS [34].