Discussion

With the advent of prostate specific antigen (PSA) screening PNBx account for an increasing part of histopathologists’ workload. However, the impact on histopathology laboratories is even greater due to the increasing number of cores taken from each patient as well as issues related to the handling of these thin needle cores. Absence of an established universal protocol for processing PNBx has led to wide variation in practice across laboratories but there is limited data documenting the degree of variation in contemporary practice.

While PNBx were taken predominantly by medical personnel such as urologists and radiologists, biopsies were not uncommonly taken by non-medical personnel, particularly in the UK. Unlike biopsies of other organs such as breast and colon, PNBx are rarely targeted and hence may be particularly suited to the latter approach. It is common in the UK for endoscopic examination and targeted colonic biopsies to be undertaken by nurses and a similar practice with respect to PNBx could become more prevalent.

Some practices such as the use of neutral buffered formalin and right/left orientation of the cores were uniformly employed by the responding laboratories but there were marked differences in other laboratory practices across Europe.

Laboratory handling of needle biopsies from other sites pose relatively little difficulty and are reasonably standardised. In contrast, PNBx pose some unique technical problems and require great attention to how they are processed as they are significantly thinner, fold easily and tend to fragment. Due to the non-targeted nature of PNBx, limited cancer in PNBx is not uncommonly associated with clinically significant cancer in the prostate gland.4 It has been shown that the detection rate of cancer in PNBx is directly proportional to the amount of tissue examined histologically.5 Suboptimal processing of PNBx could also result in under-representation of the cancer in histological sections thereby increasing the risk of equivocal diagnosis. For all these reasons optimal processing of PNBx is critical. However, it is unclear what constitutes ‘optimal’ processing as the optimal method must not only maximise the amount of tissue available for assessment but must also be cost effective.

A few studies have discussed what constitutes adequate sampling of PNBx.6–9 In the study by Lane et al,6 an average of 23.4% of total biopsy length was missed when only one level was examined compared to 7% with three levels. Brat et al7 estimated that reducing the number of levels from three to one would have resulted in 1% of adenocarcinomas being missed. However the degree of tissue loss and hence the risk of cancer being missed due to reduction in number of levels examined is likely to vary greatly between laboratories. van der Kwast et al5 suggest that examination of two levels would be adequate if the cores are embedded flat. In our survey, most laboratories routinely examined three levels of each core with almost half examining only one H&E section from each level. Since PNBx are only about 0.5 mm thick the primary reason for examining multiple levels would be to assess the entire length of the biopsy rather than to detect cancer deeper in the core.

Basal cell marker immunohistochemistry is commonly used to establish a diagnosis of cancer in PNBx. As atypical foci in PNBx are often small and ‘cut out’ in deeper levels, it has been recommended that unstained spares should be retained at the time of initial microtomy for potential immunohistochemistry.10 Retention of unstained spares was reported by only 56.1% of respondents in our survey but this practice was much more common in the UK (85.2%; p<0.0001). Cost implications of retaining unstained spares has to be weighed against the difficulty in resolving equivocal diagnosis in PNBx. Diagnostic uncertainty can often be easily resolved in organs such as the breast by targeted re-biopsy or local excision of the suspect area. However, neither is feasible in the prostate necessitating long term follow-up if re-biopsy is negative, which could result in significant patient anxiety. If unstained spare sections are not routinely retained, some small morphologically equivocal foci that may have been resolved by immunohistochemistry would remain equivocal because the focus is not represented in deeper levels. It may be cost effective to store spare sections as wax ribbons on paper rather than on specially prepared glass slides suitable for immunohistochemistry; a practice followed by 14.5% of laboratories that retained spares.

The other issue regarding unstained spare sections is how long they should be retained as this can have significant space implications especially if the cores are embedded in separate blocks. There was wide variation in how long laboratories retained spares with 41.3% of respondents retaining spares for over a year while 36.5% discarded the spares within a month of reporting the case. There is no medicolegal rationale for long term retention of unstained sections on glass slides as numerous unstained sections are routinely discarded during trimming of tissue blocks.

Only two centres in our survey routinely performed immunohistochemistry on all PNBx. While this practice could reduce the risk of missing cancers, it is expensive and would have a significant impact on laboratory and pathologist's workload. There is also the risk of overinterpreting benign glands immunonegative for basal cell markers as suspicious or even malignant. While all laboratories used at least one basal cell marker, 18.7% used p63 as the only basal cell marker. This practice could result in the rare p63 positive prostatic adenocarcinoma being misinterpreted as atrophy.11

We also noted some interesting differences between the UK and the rest of Europe in the handling of PNBx (table 6). UK laboratories more often embed multiple cores in each block and examine at least three levels of each core presumably because the cores are less often embedded flat with the whole length of the core sampled in a single level. Retention of unstained spares for potential immunohistochemistry was much more common in the UK. Use of p63 as the sole basal cell marker was more common outside the UK (26.5% vs 5.6%).

On comparing PNBx processing protocols within UK laboratories in our survey (conducted in 2010) with a previous similar UK-based survey (conducted in 2002),1 a few trends were evident. The number of cores per patient has slightly increased (median: 8 in 2002, 10 in 2010) but 30.5% of centres still obtain less than the 10 cores per patient recommended by the National Institute of Clinical Excellence guidelines in 2006.12 It is recognised that practice the UK is changing rapidly with use of 24–40 core template biopsy protocols in some centres. Use of tampers for flattening the cores during embedding that was reported by only 2% of UK laboratories in 2002 was reported by 11.4% in our survey. There was a significant increase in the number of laboratories taking unstained spare sections for potential immunohistochemistry (2002: 58%, 2010: 86.5%, p<0.0001) and a trend to retain the spares for a shorter period (<1 year: 33% in 2002, 58% in 2010; p=0.0147). Some of these changes, such as retention of spares for potential immunohistochemistry may be due to the Prostate Cancer Risk Management Programme (PCRMP) guidelines published in 2006.13 Table 7 compares the practice reported by UK laboratories with the PCRMP guidelines. It is notable that while the great majority of UK laboratories now retain spares for potential immunohistochemistry, 43.7% retain spares from less than the three levels recommended by PCRMP.

A significant limitation of our survey is that it is not possible to ascertain how representative the respondents are for pathology laboratories across Europe as there are no accessible central databases of histopathology laboratories in most countries.